SELECTING PLANT SPECIES TO OPTIMIZE WASTEWATER TREATMENT IN
CONSTRUCTED WETLANDS
by
Carrie Renee Taylor
A thesis submitted in partial fulfillment
of the requirements for the degree
of
Master of Science
in
Land Rehabilitation
MONTANA STATE UNIVERSITY
Bozeman, Montana
March 2009
©COPYRIGHT
by
Carrie Renee Taylor
2009
All Rights Reserved
ii
APPROVAL
of a thesis submitted by
Carrie Renee Taylor
This thesis has been read by each member of the thesis committee and has been
found to be satisfactory regarding content, English usage, format, citation, bibliographic
style, and consistency, and is ready for submission to the Division of Graduate Education.
Catherine A. Zabinski
Approved for the Department Land Resources and Environmental Sciences
Bruce D. Maxwell
Approved for the Division of Graduate Education
Dr. Carl A. Fox
iii
STATEMENT OF PERMISSION TO USE
In presenting this thesis in partial fulfillment of the requirements for a
master’s degree at Montana State University, I agree that the Library shall make it
available to borrowers under rules of the Library.
If I have indicated my intention to copyright this thesis by including a
copyright notice page, copying is allowable only for scholarly purposes, consistent with
“fair use” as prescribed in the U.S. Copyright Law. Requests for permission for extended
quotation from or reproduction of this thesis in whole or in parts may be granted
only by the copyright holder.
Carrie Renee Taylor
March 2009
iv
ACKNOWLEDGEMENTS
I would like to thank my advisors Cathy Zabinski, Paul Hook, and Otto Stein for
providing me the opportunity to learn more about wetland ecosystems and about the
research process. I am grateful for their ideas and help designing, analyzing and
reviewing this project. Many people across campus assisted me in this research. John
Neuman provided technical assistance with ion chromatography. He and Mark Burr’s
expertise and guidance allowed me to perform root oxygen loss measurements. Tim
McDermott’s lab provided space and assistance. Thanks to David Baumbauer for
maintaining a challenging temperature schedule for experiments in the Plant Growth
Center. Rickey Schultz helped me start my work in the greenhouse and introduced me to
water chemistry sampling methods. Melissa Schroeder and Jill Flores spent many hours
harvesting plants and analyzing roots. Finally, I would like to thank my parents who
have provided support and numerous learning experiences that have made me who I am
today.
Funding was provided by the USDA-NRI grant number 2004-35102-14832 and a
National Science Foundation GK-12 fellowship.
v
TABLE OF CONTENTS
1. INTRODUCTION ...........................................................................................................1
Background ....................................................................................................................1
Natural Wetlands .....................................................................................................1
Wetlands as Treatment Systems ..............................................................................4
Research Objectives.......................................................................................................6
2. SEASONAL PLANT EFFECTS ON WASTEWATER TREATMENT ........................8
Introduction....................................................................................................................8
Materials and Methods...................................................................................................9
Plant Materials .........................................................................................................9
Experimental Design and Operation......................................................................11
Water Sampling and Chemical Analysis ...............................................................12
Statistical Analysis.................................................................................................13
Results..........................................................................................................................14
COD Removal........................................................................................................15
Sulfate Concentration.............................................................................................26
Relationships Between COD Removal and Sulfate Concentration .......................30
Redox Potential......................................................................................................31
Discussion ....................................................................................................................32
3. PLANT TRAIT INFLUENCES ON WASTEWATER TREATMENT .......................38
Introduction..................................................................................................................38
Materials and Methods.................................................................................................42
Species Selection ...................................................................................................42
Root Oxygen Loss..................................................................................................42
Experimental Design........................................................................................42
Root Oxygen Loss Measurements ...................................................................44
Root Anatomy and Morphology ............................................................................46
Experimental Design........................................................................................46
Aerenchyma Measurement ....................................................................................46
Root Measurement .................................................................................................46
Flooding Tolerance ................................................................................................47
Statistical Analysis.................................................................................................47
Results..........................................................................................................................49
Root Oxygen Loss..................................................................................................49
Root Anatomy and Morphology ............................................................................51
Flooding Tolerance ................................................................................................52
Relationships Between Plant Traits and Plant Effects on Wastewater Treatment.54
vi
TABLE OF CONTENTS – CONTINUED
ROL and COD Removal..................................................................................54
Wetland Indicator Status, Flooding Tolerance, and COD Removal................55
Discussion ....................................................................................................................56
Relationships Between Wastewater COD Removal and ROL ..............................57
Relationships Between Wastewater COD Removal and Plant Attributes .............60
Selecting Appropriate Species for Constructed Wetlands .....................................64
4. CONCLUSIONS............................................................................................................67
Seasonal Plant Effects on Wastewater Treatment .......................................................67
Plant Selection for Constructed Wetlands ...................................................................69
REFERENCES ..................................................................................................................73
vii
LIST OF TABLES
Table
Page
1. Plant species studied and their Wetland Indicator Status .......................................10
2. COD removal on day 6 of 4˚C batches in 2007 and 2008 and the 24˚C batch in
2007........................................................................................................................17
3. Correlations between temperature and COD removal and temperature and SO 4
concentrations ........................................................................................................18
4. COD removal on day 6 of batches at all temperatures ...........................................23
5. SO 4 concentrations on day 6 of 4˚C batches in 2007 and 2008 and the 24˚C
batch in 2007..........................................................................................................27
6. SO 4 concentrations on day 6 of batches at all temperatures...................................28
7. Plant species studied and their Wetland Indicator Status .......................................43
8. Root oxygen loss at 4˚C and 24˚C ..........................................................................50
9. Percent aerenchyma below the root-shoot junction ................................................51
10. Length per root diameter class..............................................................................52
11. Plant species studied and their attributes and influences on wastewater ..............72
viii
LIST OF FIGURES
Figure
Page
1. COD removal, SO 4 concentration, and Eh over 20-day batch incubations at
4˚C in 2007 and 24˚C in 2007................................................................................16
2. Seasonal variation in day-6 COD removal, SO 4 concentration, and Eh for
Cyperaceae ............................................................................................................19
3. Seasonal variation in day-6 COD removal, SO 4 concentration, and Eh for
Juncaceae...............................................................................................................20
4. Seasonal variation in day-6 COD removal, SO 4 concentration, and Eh for
Poaceae..................................................................................................................21
5. Seasonal variation in day-6 COD removal, SO 4 concentration, and Eh for
Iridaceae, Lamiaceae, and Typhaceae...................................................................22
6. COD removal and SO 4 concentration for each species and unplanted control
at 4˚C and 24˚C......................................................................................................25
7. Relationships between COD removal and SO 4 concentrations at 4˚C and 24˚C ..30
8. Root cross sections of C. canadensis and D. cespitosa ..........................................52
9. Biomass of plant species when flooded divided by biomass when drained ...........53
10. Correlation between COD removal and root oxygen loss at 4˚C and 24˚C..........55
ix
ABSTRACT
Constructed wetlands are used around the world for treating domestic,
agricultural, and industrial wastewater, stormwater runoff, and acid mine drainage.
Plants may affect efficacy of wastewater treatment through their influence on microbial
activity by creating attachment sites and releasing carbon exudates and oxygen. My
research investigated seasonal plant effects on wastewater treatment by monitoring water
chemistry in model subsurface wetlands planted with monocultures of 19 plant species
and unplanted controls. Chemical oxygen demand (COD) removal, an indicator of water
quality, declined during colder temperatures in the unplanted control, likely caused by a
decrease in microbial activity. In contrast, wetlands with most plant species had constant
COD removal across seasons. Redox potential and sulfate concentrations were measured
as indirect measurements of the oxygenation of the wastewater. Wetlands that had a
decline in COD removal during cold temperatures had constant low redox potential and
sulfate concentrations throughout the seasons. Wetlands with high COD removal across
seasons had elevated redox potentials and sulfate concentrations during the winter,
indicating elevated oxygen availability, which may offset the negative temperature effect
on microbial processes. I measured root oxygen loss (ROL) in the summer and the
winter to determine whether oxygen release was sufficient to influence wastewater
treatment and cause seasonal and species-specific effects on water chemistry. COD
removal and ROL were positively correlated at 4˚C but not at 24˚C; however, the amount
of root oxygen release only accounted for a portion of the required oxygen to facilitate
plant’s influence on COD removal. Flooding tolerance was quantified for each species
by comparing plants’ biomass between flooded and drained conditions. Plants’ botanical
grouping, Wetland Indicator Status, and flooding tolerance were compared to plants’
influences on wastewater treatment to determine whether easily measured plant traits can
be used to identify plants that will optimize wastewater treatment. All the sedges and
rushes, obligate wetlands species, and 8 of 9 flood-tolerant plants had greater COD
removal than the control at 4˚C, the coldest temperature incubation. These results can be
applied for wetland design by selecting plant species to optimize wastewater treatment,
especially in cold climates.
1
INTRODUCTION
Background
Natural Wetlands
Wetlands are diverse ecosystems teeming with life and activity which can activate
all of our senses. After the disappearance of more than 50% of the wetlands in the U.S.,
these habitats have been increasingly recognized for their unique ecological functions and
are protected by the federal government. Wetlands are defined by the US Army Corps of
Engineers (Federal Register 1982) and the Environmental Protection Agency (Federal
Register 1980) as “areas that are inundated or saturated by surface or ground water at a
frequency and duration sufficient to support, and that under normal circumstances do
support, a prevalence of vegetation typically adapted for life in saturated soil conditions.”
A great diversity of ecosystems around the earth are considered wetlands, including salt
marshes, tidal freshwater marshes, inland marshes, riverine marshes, forested wetlands,
mangrove swamps, and peatlands.
These diverse ecosystems perform similar landscape functions, which can be
divided into three main categories: hydrology, habitat, and biogeochemistry (Cronk and
Fennessy 2001). Hydrologic functions include regulating flows within watersheds and
reducing flood and storm damage since wetlands act as water storage basins. Wetlands
are among the most productive ecosystems in the world and provide habitat to a diversity
of microbes, plants, insects, amphibians, reptiles, birds, fish, and mammals. A number of
biogeochemical processes occur in wetlands due to saturation and high primary
2
productivity. Biogeochemical processes transport and transform chemicals through
settling, absorption, microbial processes, plant uptake, and other processes, are part of
global water and nutrient cycles and can remove many common pollutants that enter
wetlands from surface and groundwater.
Water saturation in wetlands largely determines how the soil develops and the
types of plant communities present. When saturated, soil pore spaces are filled with water
rather than gas, reducing the rate of oxygen diffusion by a factor of 10,000 (Cronk and
Fennessy 2001). Slow oxygen diffusion coupled with the respiration demands of plant
roots and soil microorganisms result in anaerobic conditions. Because of this, reduced
inorganic and organic compounds accumulate in the soil. Low oxygen levels and
reduced compounds create a stressful environment for plant growth. Only those plants
that are physiologically and structurally adapted to these anoxic conditions are capable of
growing in wetlands.
Wetland plants have various traits and mechanisms for avoiding or tolerating
anoxia. Flood-tolerant species’ metabolic response to anaerobic conditions differs from
less flood-tolerant plants (McManmon and Crawford 1971, Drew 1997, Schluter and
Crawford 2001, Crawford 2003). Flooding-sensitive plants are severely damaged by 24
hours of anoxia. Some plants have strategies to withstand temporary anoxia by
accelerating anaerobic metabolism, while other plants are well adapted to survive and
grow for extended periods in anaerobic conditions. Wetland plants which are flood
tolerant may down-regulate metabolism and have adapted structurally to create an
aerobic environment within anoxic conditions. Aerenchyma tissue, which has gas-filled
3
spaces, called lacunae, separating cells, provides a low-resistance internal pathway for the
exchange of gases between the plant parts above the water and the submerged tissues
(Armstrong 1964, Armstrong 1979, Justin and Armstrong 1987).
Oxygen is
transported to roots, and is either used in respiration or released to the surrounding soil or
water matrix (Armstrong 1971a). Many wetland plants prevent excessive oxygen loss
from the basal root zones by forming roots with a complete or partial ‘barrier’ to radial
oxygen loss in their epidermis, exodermis or subepidermal layers with lignified or
suberized cells (Armstrong 1979, Jackson and Drew 1984, Armstrong and Armstrong
1988, Jackson and Armstrong 1999, Visser et al. 2000b).
Root oxygen loss to the surrounding soil or water matrix results in the oxidation
and detoxification of potentially harmful reducing substances in the rhizosphere thus
benefitting plants (Armstrong 1979, Barko and Smart 1983, Tolley et al. 1986, Ernst
1990, Koncalova 1990). Reduced manganese, iron, and sulfur interfere with enzyme
structure, chlorophyll formation, and metabolism, respectively (Cronk and Fennessy
2001). The reduced substances are products of microorganisms’ anaerobic metabolism.
Under saturated soil conditions, anaerobic soil microorganisms use terminal electron
acceptors other than oxygen, reducing nitrate (NO 3 -), manganese ions (Mn4+), ferric ions
(Fe3+), sulfate (SO 4 2-), and carbon dioxide (CO 2 ). Root oxygen loss provides an
energetically favorable terminal electron acceptor for microbial respiration resulting in
oxidation of organic carbon, sulfide (S2-), and reduced iron and manganese.
4
Wetlands as Treatment Systems
Constructed wetlands designed specifically for wastewater treatment take
advantage of purifying processes found in natural wetlands. They can provide an energyefficient, cost-effective, and low maintenance alternative to conventional wastewater
treatment technologies (Kadlec and Knight 1996). They also have ancillary benefits that
include creating habitat for plants and wildlife and an aesthetic greenspace in
communities. Constructed wetlands are used around the world to treat domestic,
agricultural, and industrial wastewater, stormwater runoff, and acid mine drainage. They
are designed either as free water surface wetlands with standing water or as subsurface
flow wetlands with water below the soil or gravel surface. Wastewater is delivered in
batches or continuously and flows either horizontally or vertically through the substrate.
Wastewater in constructed wetlands is treated by physical, chemical, and
biological interactions between wastewater, wetland media, microorganisms, and plants.
As in natural wetlands, basic treatment mechanisms include sedimentation, chemical
precipitation, adsorption, uptake by vegetation, and microbial processes. Microbial
activity has been considered to be the primary mechanism for removing most pollutants
including organic carbon and nitrogen (Hatano et al. 1993, Kadlec and Knight 1996).
Constructed wetlands are designed to increase the predictability and efficiency of
treatment, enabling use of wetlands that are less land-intensive.
Plants are major components of wetlands, yet relatively little is known about their
effects on wastewater and microbial processes or seasonal responses to temperature. A
greater understanding of plants’ seasonal influences would allow designers to manipulate
5
another variable to increase the efficiency of wastewater treatment. In currant practice,
relatively few plant species are used in constructed wetlands. Of the ‘first generation’ of
constructed wetlands, one-third of the free water surface systems used only Typha spp.
and about 40% of the subsurface systems used only Scirpus spp. (Reed and Brown 1992).
Phragmites australis is the most widely used plant species in European systems (USEPA
1993). The low number of species used is likely due to the lack of knowledge regarding
plant influences on wastewater treatment and a shortage of evidence for species-specific
effects.
Past studies that compared the inflow and outflow chemistry in constructed
wetlands either with or without plants have revealed contradictory results. Some
publications report that plants have little direct impact on wastewater treatment (Brix and
Schierup 1990, Haberl and Perfler 1990, Kadlec and Knight 1996, USEPA 2000, Vacca
et al. 2005), while others studies have found improved treatment with plants present
(Coleman et al. 2001, Hunter et al. 2001, Lin et al. 2002, Picard et al. 2005,
Iamchaturapatr et al. 2007, Yang et al. 2007). These contradictions may reflect the plants
selected to test against unplanted wetlands or the season in which studies have been
conducted.
In addition to determining that the presence of plants improved treatment over
unplanted wetlands, a few studies have documented species-specific effects. However,
the ranking of species performance differed. Gersberg et al. (1986), Allen et al. (2002),
Fraser et al. (2004), and Stein and Hook (2005) found that Scirpus spp. and Carex spp.
had greater nutrient removal than Typha spp., while Akratos and Tsihrintizis (2007)
6
found that Typha spp. had the greatest removal. No direct measurements were taken to
determine the mechanisms behind species’ influences in these studies, but researchers
conjectured that differences were due to either root oxygen loss or direct plant uptake.
Plants can improve treatment via the direct uptake of nutrients, but plant nutrient
content cannot account for the enhanced nutrient removal from wastewater (Gersberg et
al. 1986, Bachand and Horne 2000, Lin et al. 2002, Akratos and Tsihrintizis 2007).
Rather, plant’s influences on microbial activity, which include providing attachment sites
and releasing carbon exudates and oxygen, are believed to be their primary role in
treatment wetlands (Reed and Brown 1992, Brix 1997, Tanner 2001).
Research Objectives
The purpose of my research was to determine whether plants affect treatment
efficacy in constructed wetlands, and if so, by what means. Additionally, my research
evaluated species-specific effects and investigated relationships between plant traits and
influences on wastewater treatment, so that plant traits can guide species selection to help
optimize wastewater treatment. The objectives of the experiment in Chapter 2 were to
investigate effects of plants and season on wastewater treatment. Chapter 2 describes an
experiment which measured and compared water chemistry (COD removal, sulfate
concentrations, and redox potential) across seasons and compared unplanted columns and
columns planted with monocultures of 19 plant species.
The third chapter addresses the mechanisms by which plants increase treatment
efficacy. The objectives of the experiments included in this chapter were to determine
7
whether root oxygen loss is a mechanism driving plants’ seasonal effects, and whether
relationships exist between plant traits and their influences on wastewater. This could be
valuable for increasing a designer’s ability to select species that maximize wetland
efficacy. I measured root oxygen loss in the summer and the winter to determine whether
oxygen release was sufficient to influence wastewater treatment and cause the seasonal
and species-specific effects on water chemistry documented in Chapter 2. To determine
whether there is a relationship between plant traits and treatment efficacy, COD removal
data reported in Chapter 2 were compared to the plants’ Wetland Indicator Status,
flooding tolerance, root morphology, and root anatomy.
Together, these studies create a bridge between the biology and ecology of
wetland plants and applied interests in engineering wastewater treatment systems.
Increasing our knowledge of the interactions in the ‘black box’ of the constructed
wetland substrate could help provide a more rational basis for plant selection and increase
wastewater treatment efficacy. In addition, ancillary functions of artificial wetlands, such
as providing a diverse habitat would benefit from the increased use of native plants are
shown to provide good water treatment (USEPA 2000).
8
SEASONAL PLANT EFFECTS ON WASTEWATER TREATMENT
Introduction
Constructed wetlands can provide an energy-efficient and cost-effective
alternative to conventional wastewater treatment technologies, but their efficacy in cold
climates was uncertain for many years. Numerous case studies during the last decade
have now demonstrated effective wastewater treatment in subsurface flow wetlands in
cold regions (Mander and Jenssen 2003). Nonetheless, understanding of the factors that
allow effective treatment year-round, and how they can be optimized, remains poor.
Allen et al. (2002) reported evidence that plants can contribute to effective COD removal
at low temperatures.
The role of plants in constructed wetlands has long been debated. Some studies
have found that plants improve wastewater treatment (Gersberg et al. 1986, Reddy et al.
1989, Lin et al. 2002, Fraser et al. 2004, Picard et al. 2005, Stein and Hook 2005, Akratos
et al. 2007, Yang et al. 2007), while others have found little or no benefit (Brix and
Schierup 1990, Haberl and Perfler 1990, Vacca et al. 2005). In practice, only a handful
of plants, including Phragmites australis and Typha, Scirpus and Schoenoplectus species,
are widely used in constructed wetlands. Most research has focused on these species and
their effects in mild climates.
Plant effects during cold periods have received little attention. Plants’ primary
role in treatment wetlands is thought to be their influence on microbial activity through
creating attachment sites and releasing carbon exudates and oxygen (Reed and Brown
9
1992, Tanner 2001). While variation in temperature affects microbial activity directly,
seasonal cycles of plant growth and physiology may also influence ROL, other root-zone
processes and, consequently, seasonal wastewater treatment. Because morphological,
physiological, and phenological characteristics vary widely among wetland plants,
different species’ effects on seasonal CW performance may also vary. Hook et al. (2003)
found that seasonal patterns of COD removal differed among three species and
hypothesized that this resulted from differences in ROL, particularly at low temperatures.
The purpose of this study was to compare the effects of a large number of
species on seasonal removal of organic carbon from wastewater. Variation in wastewater
treatment was evaluated at temperatures from 4˚C to 24˚C by measuring COD, sulfate,
and redox potential in batch-loaded model wetlands planted with monocultures of 19
species as well as unplanted controls. Specific questions addressed were:
1. Does the presence of plants affect seasonal wastewater chemistry and are
those effects common for many species or species- specific?
2. Is variation in COD removal with different plant species and at different
temperatures related to two indicators of root-zone oxidation status, sulfate
and redox potential?
Materials and Methods
Plant Materials
A diverse selection of sedges, rushes, grasses, and forbs was chosen for the study
(Table 1). Plants ranged from Facultative to Obligate Wetlands species (U.S. Fish and
10
Wildlife Service 1988), represented several families, and included plants in common or
limited use in constructed wetlands or wetland restoration. Selection was also based on
their availability and use in past constructed wetland research. The plants were either
purchased from nurseries or transplanted from the field.
Table 1: Plant species studied and their Wetland Indicator Status. Obligate Wetland
species occur in wetlands >99% of the time, Facultative Wetland species 67-99% of the
time, and Facultative species 34-66% of the time.
Family
Species
Common name
Cyperaceae
Typhaceae
Cyperaceae
Juncaceae
Poaceae
Iridaceae
Cyperaceae
Poaceae
Lamiaceae
Obligate Wetland Species
Carex aquatilis Wahlenb.
Carex bebbii Olney ex Fernald
Carex nebrascensis Dewey
Carex utriculata Boott
Schoenoplectus acutus (Muhl. ex Bigelow)
A. Love & D. Love
Typha latifolia L.
Facultative Wetland Species
Carex praegracilis W. Boott
Juncus arcticus Willd.
Juncus torreyi Coville
Calamagrostis canadensis (Michx.) P.
Beauv.
Deschampsia cespitosa (L.) P. Beauv.
Phalaris arundinacea L.
Phragmites australis (Cav.) Trin. ex Steud.
Iris missouriensis Nutt.
Facultative Species
Carex microptera Mack.
Hordeum jubatum L.
Leymus cinereus (Scribn. & Merr.) A. Love
Panicum virgatum L.
Prunella vulgaris L.
water sedge
Bebb’s sedge
Nebraska sedge
Northwest Territory sedge
hardstem bulrush
broadleaf cattail
clustered field sedge
arctic rush
Torrey’s rush
bluejoint
tufted hairgrass
reed canarygrass
common reed
Rocky Mountain iris
smallwing sedge
foxtail barley
basin wildrye
switchgrass
common selfheal
11
Experimental Design and Operation
The greenhouse experiment was conducted in the Plant Growth Center at
Montana State University in Bozeman, MT (46˚N, 111˚W). Six replicates of each
species and controls were planted in model subsurface wetlands (“columns”) in a
randomized block design. Each block contained 20 model subsurface wetlands
consisting of 15 cm diameter by 30 cm tall polyvinyl chloride (PVC) columns planted
with one of the 19 plant species in monoculture or left unplanted as a control. Plant
material was either nursery grown or collected from the field. Roots were washed free of
soil and cut to approximately 15 cm in length and planted in 1-5 mm gravel.
During an establishment period of six months, plants were grown at 16˚C with
supplemental lighting (GE Multi-Vapor MVR1000/C/U) to maintain a 16-hour day
length. From June 20, 2006, through February 12, 2008, greenhouse temperature was
changed every 60 days to mimic natural seasonal temperature cycles. The temperature
sequence was 24, 16, 8, 4, 8, 16, 24, 16, 8, and 4˚C. Supplemental lighting was not used
so that patterns of natural light and controlled temperature would induce seasonal cycles
of plant dormancy and growth.
A synthetic wastewater simulating secondary domestic effluent was used
throughout the experiment. The wastewater was made with 0.58 mM sucrose
(C 12 H 22 O 11 ), 0.73 mM Primatone (hydrolyzed meat protein, Sigma Chemical Company),
10.7 mM NH 4 Cl, 0.25 K 2 HPO 4, 0.25 mM MgSO 4 , 0.16 mM H 3 BO 3 , 0.05 mM MnSO 4 ,
0.03 mM ZnSO 4 , 0.02 mM Na 2 MoO 4 , 0.01 mM CaCl 2 , 0.01 mM KI, 3.2 µM CuSO 4 ,
and 1.0 µM FeCl 3 . Throughout the experiment, the wastewater influent was sampled
12
immediately before being added to the experimental columns and had mean and standard
error of 490 ± 4.3 mg/L COD, 0.8 ± 0.1 mg/L NO 3 , 8 ± 0.3 mg/L PO 4 , and 14 ± 0.5
mg/L SO 4 . To simulate batch operation in subsurface flow wetlands, columns were
drained and filled with synthetic wastewater every 20 days during the plants’
establishment period and throughout the experiment. Three wastewater batches occurred
within each 60-day temperature period. Columns had drains at the base and were
connected to a standpipe to maintain the water level just below the gravel surface.
Water Sampling and Chemical Analysis
Chemical oxygen demand (COD), sulfate (SO 4 ) concentration, and redox
potential (Eh) were measured to determine whether plant species or seasons affected
wastewater treatment (COD removal) and aspects of water chemistry related to oxidation
(SO 4 and Eh). Sampling began in July 2006 and ended in February 2008.
In the center of each column, a sampling tube (0.3 cm inner-diameter vinyl
tubing) with an intake 15 cm below the gravel surface was attached to a vertical 13-cm
ID CPVC pipe. Water samples were collected with a 20 mL syringe during the third 20day wastewater batch of each temperature period. Collection from 3 replicates of each
species and unplanted control occurred on days 0, 1, 3, 6, 9 and 20 during 24˚C and 4˚C
batches, and on days 1 and 6 during the 16˚C and 8˚C batches. Influent wastewater was
collected immediately after wastewater was made and prior to filling the experimental
columns. Samples of the influent wastewater and bulk solution from the columns were
analyzed immediately for COD. COD was measured colorimetrically using HACH
reagents and spectrophotometer (0-1500 mg/L COD range test using dichromate
13
reagents, DR 2010 Spectrophotometer, HACH Company, Loveland, CO). Additional
samples were filtered (0.2 µm filter) to sterilize, stored in sterile test tubes at 2˚C, and
then analyzed for SO 4 using ion chromatography (Dionex Corporation, Sunnyvale, CA).
Platinum redox electrodes (Faulkner et al. 1989) were permanently installed at 15
cm below the gravel surface in two replicates of each plant species. Columns with redox
probes were connected by a salt bridge with a centrally located saturated calomel
reference electrode (Veneman and Pickering 1983). Redox potential was measured
automatically every 4 hours with a computer and multiplexer. Redox potential was
calculated from measured electrode potential by adding a 250 mV as the correction factor
(Vepraskas et al. 2002).
Statistical Analysis
To establish whether plant species and seasons had an effect on wastewater
treatment, COD removal, SO 4 concentration, and Eh were compared among planted and
control columns across seasons. Values measured on day 6 of each wastewater batch
were used in all statistical analysis since this is a reasonable duration for batch operations
in constructed wetlands. Additionally, patterns among plant species on day 6 of the batch
were typically similar to those found on other days of the incubation.
Analysis of variance (ANOVA, SPSS, Inc., Chicago, IL Version 15.0) was used
to compare plant species effects on COD and SO 4 within wastewater batches. Planned
contrasts were conducted to test differences between each species and the unplanted
control. Post hoc Tukey HSD (honestly significant difference) tests were used to identify
14
differences among plant species. The significance level was set at α = 0.05. Redox
potentials were not compared statistically because only two replicates were monitored.
Seasonal temperature effects on COD and SO 4 were analyzed in two ways.
Repeated measures analysis of variance (ANOVAR, SPSS, Inc., Chicago, IL Version
15.0) was used to compare 4 and 24˚C batches, which represent the extremes of
temperatures tested. Plant species and the control were analyzed individually. The
significance level was set at α = 0.05. Additionally, Pearson’s correlation (SPSS, Inc.,
Chicago, IL Version 15.0), using data from across the seasons, was used to determine
COD removal and SO 4 concentration response to temperature for each plant treatment
and the unplanted control. Measurements from the first sampling period (24˚C in 2006)
were not used in ANOVARs and correlations so as to exclude effects associated with the
start-up of the experiment; COD removal was markedly lower in the first batch, and
patterns of plant effects were not consistent with those seen throughout the rest of the
study. Regression (SPSS, Inc., Chicago, IL Version 15.0) was used to determine
relationships between COD removal and SO 4 during each temperature period.
Results
COD Removal
COD removal throughout wastewater batches was influenced by plant species but
showed limited differences between summer and winter. Rapid COD removal occurred
within 24 hours in both planted and unplanted columns during 20-day wastewater batches
at all temperatures (Figure 1). In some cases the column’s maximum removal was
15
reached within 24 hours, and in others removal continued for up to 20 days; maximum
removal was normally achieved by day 6. COD removal on day 6 ranged from
45% to 100%, with a median value of 89% for planted columns and 66% for unplanted
columns. When present, differences among plant species emerged by day 3 and normally
persisted throughout the batch.
Figure 1 shows controls and a subset of the 19 species that represent the range of
removal patterns seen over the course of the warmest and coldest batches. The unplanted
control was relatively effective at COD removal during warm temperatures, reaching
85% and 93% removal by day 20 of the 24˚C 2006 and 2007 batches, respectively;
removal in controls was less effective during cold temperatures, with 62% and 64%
removal during the 4˚C 2007 and 2008 batches. In columns planted with many of the
species, COD removal was nearly complete within 3 days in both summer and winter. C.
bebbii, S. acutus, and D. cespitosa are examples of this pattern. P. virgatum is an
example of a species with lower removal at 4˚C than 24˚C, and T. latifolia is
representative of the plants that had intermediate patterns of COD removal.
Seasonal differences in removal measured on day 6 were statistically significant
for a minority of species. Repeated measures ANOVA comparing results from 4˚C and
24˚C indicated that controls and columns with I. missouriensis, L. cinereus and P.
virgatum had significantly lower COD removal at 4˚C, while C. utriculata and P.
arundinacea had significantly higher removal at 4˚C (Table 2). Other species did not
differ significantly between 4 and 24˚C or showed inconsistent differences. Only the
controls and three plant species showed significant correlations between COD removal
16
COD removal (%)
4˚C 2007
24˚C 2007
100
80
60
40
20
0
Control
C. bebbii
S. acutus
D. cespitosa
P. virgatum
T. latifolia
SO4 (mg/L)
15
10
5
0
Eh (mV)
600
Control
C. bebbii
S. acutus
D. cespitosa
P. virgatum
T. latifolia
400
200
Control
Control
C. bebbii
C. bebbii
S. acutus S. acutus
D. cespitosa D. cespitosa
P. virgatum P. virgatum
T. latifolia T. latifolia
0
-200
0
3
6
9
Day
20 0
3
6
9
Day
20
Figure 1: COD removal (%), SO 4 concentration (mg/L), and Eh (mV) over 20-day batch
incubations at 4˚C in 2007 and 24˚C in 2007. Symbols in COD removal and SO 4 graphs
represent the mean of three replicates. Error bars represent ± one standard error. Day
zero represents data collected immediately after filling the columns with wastewater. See
Tables 3-6 for means, standard errors, and statistical comparisons for COD and SO 4 .
Lines in Eh graphs represent the mean of two replicates with measurements taken at
4-hour intervals.
17
and temperature when using data from all temperature incubations (Table 3).
Temperature had a positive effect on COD removal only in the unplanted controls and
columns with I. missouriensis, which had little biomass. Removal increased at lower
temperatures in columns planted with C. nebrascensis and C. utriculata.
Table 2: COD removal (%) on day 6 of 4˚C batches in 2007 and 2008 and the 24˚C batch
in 2007. COD values are means of three replicates ± one standard error. Values with an
asterisk (*) are significantly different from other values in the same row (p<0.05,
Repeated Measures ANOVA).
4˚C 2007
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. arcticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
58 ± 1
90 ± 3
98 ± 1
83 ± 4
96 ± 2
96 ± 2
92 ± 2
89 ± 4
94 ± 2
73 ± 11
99 ± 1
71 ± 3
69 ± 2
66 ± 3
87 ± 1
67 ± 5
67 ± 5
87 ± 1
83 ± 6
*
24˚C 2007
88 ± 2
98 ± 0
100 ± 0
82 ± 3
93 ± 3
97 ± 3
83 ± 1
94 ± 2
96 ± 4
92 ± 6
94 ± 4
96 ± 2
86 ± 4
79 ± 2
92 ± 2
79 ± 4
67 ± 6
92 ± 1
77 ± 8
89 ± 4
*
*
*
*
*
*
4˚C 2008
66 ± 2
97 ± 1
98 ± 1
96 ± 3
97 ± 1
97 ± 1
99 ± 1
99 ± 0
96 ± 2
91 ± 2
85 ± 2
98 ± 1
79 ± 1
72 ± 1
71 ± 3
92 ± 4
89 ± 1
67 ± 3
74 ± 2
92 ± 2
*
COD removal was consistently high across all seasons with five of the six sedges
(C. aquatilis, C. bebbii, C. nebrascensis, C. praegracilis, and C. utriculata, 83-100%
removal), both rushes (J. arcticus and J. torreyi, 89-97%), S. acutus, (87-100%), and D.
cespitosa (95-100%) (Figures 2-5, Table 4). Removal was more variable for other
species (53-94%) with lower average removal across seasons (70-88%). Plant species
18
rankings for COD removal were similar across batches. Species with the highest removal
at cold temperatures also had the highest removal at warm temperatures (Figure 6).
Table 3: Correlations between temperature and COD removal (%) and temperature and
SO 4 concentrations (mg/L). COD removal and SO 4 concentration values are from the
sixth day of each wastewater batch. Analyses include data for batches from 16˚C in 2006
through 4˚C in 2008. The coefficient of correlation is significantly different than 0 if
followed by an asterisk (*).
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. arcticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
COD removal
and Temperature
r
+ 0.77 *
+ 0.37
- 0.20
- 0.55
- 0.67 *
- 0.26
- 0.93 *
- 0.20
+ 0.32
- 0.37
+ 0.49
- 0.40
+ 0.51
+ 0.40
+ 0.60
- 0.40
- 0.24
+ 0.67 *
- 0.45
- 0.10
SO 4 and
Temperature
r
+ 0.35
- 0.73
- 0.84
- 0.63
- 0.79
- 0.77
- 0.85
- 0.77
- 0.77
- 0.79
- 0.14
- 0.78
- 0.33
- 0.32
+ 0.40
- 0.69
- 0.48
+ 0.05
- 0.25
- 0.30
*
*
*
*
*
*
*
*
*
*
*
Seasonal differences in plant effects are also apparent in statistical comparison
among plants and controls during individual batches (Table 4). There were few
significant differences between plant species and the unplanted control at 24 and 16˚C in
2006 and 24˚C in 2007; the species that differed from controls were not consistent for
these batches. In contrast, 13-18 species differed from controls during the other batches;
in every case, removal was greater in planted columns than controls.
COD removal (%)
19
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
100
80
60
40
SO4 (mg/L)
15
10
5
600
400
200
0
-200
-400
24
o
Temperature ( C)
Eh (mV)
0
16
Ju
ly
Se -06
pt
N 06
ov
J a -06
M n-0
ar 7
ch
M -07
ay
Ju - 07
ly
Se -07
pt
N -0 7
ov
Ja 0 7
n08
8
4
Figure 2: Seasonal variation in day-6 COD removal (%), SO 4 concentration (mg/L), and
Eh (mV) for Cyperaceae. SO 4 and COD removal symbols represent the mean of three
replicates. Eh symbols represent the mean of two replicates. Error bars represent ± one
standard error. See Tables 2, 4, 5, and 6 for means, standard errors, and statistical
comparisons.
COD removal (%)
20
100
Control
J. articus
J. torreyi
80
60
40
SO4 (mg/L)
15
10
5
600
400
200
0
-200
-400
24
16
Se -06
pt
N 06
ov
Ja -06
M n-0
ar 7
ch
M -07
ay
Ju -07
ly
Se -0 7
pt
N -0 7
ov
Ja 07
n08
8
4
Ju
ly
Temperature (oC)
Eh (mV)
0
Figure 3: Seasonal variation in day-6 COD removal (%), SO 4 concentration (mg/L),
and Eh (mV) for Juncaceae. SO 4 and COD removal symbols represent the mean of
three replicates. Eh symbols represent the mean of two replicates. Error bars
represent ± one standard error. See Tables 2, 4, 5, and 6 for means, standard errors,
and statistical comparisons.
COD removal (%)
21
Control
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
100
80
60
40
SO4 (mg/L)
15
10
5
0
400
200
0
-200
-400
24
16
Se 06
pt
N 06
ov
Ja -06
M n-0
ar 7
ch
M -07
ay
Ju - 0 7
ly
Se -07
pt
N -0 7
ov
Ja 07
n08
8
4
Ju
ly
-
Temperature (oC)
Eh (mV)
600
Figure 4: Seasonal variation in day-6 COD removal (%), SO 4 concentration (mg/L),
and Eh (mV) for Poaceae. SO 4 and COD removal symbols represent the mean of
three replicates. Eh symbols represent the mean of two replicates. Error bars
represent ± one standard error. See Tables 2, 4, 5, and 6 for means, standard errors,
and statistical comparisons.
COD removal (%)
22
Control
I. missouriensis
P. vulgaris
T. latifolia
100
80
60
40
SO4 (mg/L)
15
10
5
600
400
200
0
-200
-400
24
o
Temperature ( C)
Eh (mV)
0
16
Ju
ly
Se - 06
pt
N 06
ov
Ja - 0 6
M n-0
ar 7
ch
M -07
ay
J u -07
ly
Se -07
pt
N -07
ov
Ja 07
n08
8
4
Figure 5: Seasonal variation in day-6 COD removal (%), SO 4 concentration (mg/L),
and Eh (mV) for Iridaceae, Lamiaceae, and Typhaceae. SO 4 and COD removal
symbols represent the mean of three replicates. Eh symbols represent the mean of
two replicates. Error bars represent ± one standard error. See Tables 2, 4, 5, and 6
for means, standard errors, and statistical comparisons.
Table 4: COD removal (%) on day 6 of batches at all temperatures. COD values (% removal) are means of three replicates ± one
standard error. Within each column, values in bold font are significantly different than the control (p<0.05, planned contrasts).
Values in a column with the same letter are not significantly different from each other (p≥0.05, Tukey HSD).
16˚C 2006
92 ± 1 cd
89 ± 1 cd
90 ± 2 cd
80 ± 6 bcd
91 ± 2 cd
96 ± 1 d
93 ± 2 d
87 ± 3 bcd
89 ± 0 cd
94 ± 1 d
90 ± 1 cd
95 ± 2 d
76 ± 1 bcd
66 ± 5 ab
70 ± 6 abc
92 ± 1 cd
53 ± 2 a
83 ± 6 bcd
81 ± 11 bcd
79 ± 6 bcd
84 ± 3
8˚C 2006
62 ± 4 ab
88 ± 7 cdef
98 ± 1 f
88 ± 8 bcdef
100 ± 0 f
95 ± 2 f
97 ± 1 f
92 ± 5 ef
90 ± 6 def
95 ± 3 f
65 ± 6 abcd
99 ± 1 f
66 ± 5 abcd
62 ± 5 a
69 ± 4 abcde
81 ± 4 abcdef
65 ± 5 abcd
63 ± 4 abc
87 ± 6 abcdef
81 ± 6 abcdef
83 ± 3
4˚C 2007
58 ± 1 a
90 ± 3 cde
98 ± 1 e
83 ± 4 bcde
96 ± 2 e
96 ± 2
92 ± 2
89 ± 4
94 ± 2
73 ± 11
99 ± 1
71 ± 3
69 ± 2
66 ± 3
87 ± 1
67 ± 5
67 ± 5
87 ± 1
83 ± 6
84 ± 3
e
de
bcde
de
abcd
e
abcd
abc
ab
bcde
ab
ab
bcde
bcde
8 ˚C 2007
65 ± 9 ab
95 ± 2 def
98 ± 0 def
93 ± 3 def
98 ± 1 ef
97 ± 1 def
99 ± 1 f
94 ± 2 def
96 ± 0 def
96 ± 0 def
80 ± 4 bcd
98 ± 1 ef
88 ± 4 cdef
61 ± 3 a
81 ± 5 bcde
89 ± 1 cdef
73 ± 3 abc
83 ± 2 cdef
84 ± 4 cdef
83 ± 3 cdef
89 ± 2
16˚C 2007
66 ± 1 a
98 ± 0 e
99 ± 0 e
92 ± 4 de
96 ± 3 e
95 ± 2 e
90 ± 5 de
97 ± 1 e
97 ± 2 e
90 ± 2 de
69 ± 2 abc
100 ± 0 e
88 ± 7 cde
69 ± 3 ab
75 ± 4 abcd
65 ± 3 a
88 ± 5 cde
70 ± 4 abc
62 ± 5 a
85 ± 4 bcde
86 ± 3
23
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. arcticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
All species
24˚C 2006
66 ± 4 abc
57 ± 9 ab
91 ± 2 c
72 ± 1 abc
70 ± 7 abc
86 ± 1 bc
77 ± 3 bc
80 ± 3 bc
88 ± 4 c
82 ± 3 bc
45 ± 16 a
75 ± 13 abc
78 ± 3 bc
76 ± 4 abc
69 ± 4 abc
71 ± 7 abc
74 ± 5 abc
84 ± 5 bc
94 ± 0 c
75 ± 3 abc
76 ± 3
Table 4 continued: COD removal (%) on day 6 of batches at all temperatures. COD values (% removal) are means of three replicates
± one standard error. Within each column, values in bold font are significantly different than the control (p<0.05, planned contrasts).
Values in a column with the same letter are not significantly different from each other (p≥0.05, Tukey HSD). Averages in the righthand column are for batches 16˚C 2006 through 4˚C 2008; the first batch reflected start-up effects and is excluded. Averages were not
compared statistically.
16˚C 2007
69 ± 2 a
95 ± 0 e
91 ± 2 de
82 ± 4 abcde
95 ± 0 e
94 ± 1 e
91 ± 3 de
96 ± 1 e
94 ± 0 e
93 ± 2 e
82 ± 2 abcde
95 ± 1 e
87 ± 5 bcde
74 ± 1 ab
77 ± 6 abc
90 ± 4 cde
79 ± 2 abcd
69 ± 2 a
75 ± 1 ab
84 ± 1 bcde
86 ± 2
8˚C 2007
64 ± 1 a
93 ± 1 fghi
99 ± 0 hi
93 ± 3 fghi
99 ± 0 hi
98 ± 2 hi
97 ± 1 hi
100 ± 0 i
91 ± 3 efghi
97 ± 2 hi
78 ± 2 bcd
95 ± 1 ghi
80 ± 4 bcde
77 ± 3 bc
83 ± 3 cdefg
88 ± 2 cdefgh
87 ± 4 cdefgh
69 ± 0 ab
82 ± 2 cdef
90 ± 2 defghi
89 ± 2
4˚C 2008
66 ± 2 a
97 ± 1 e
98 ± 1 e
96 ± 3 e
97 ± 1 e
97 ± 1 e
99 ± 1 e
99 ± 0 e
96 ± 2 e
91 ± 2 de
85 ± 2 cd
98 ± 1 e
79 ± 1 bc
72 ± 1 ab
71 ± 3 ab
92 ± 4 de
89 ± 1 de
67 ± 3 a
74 ± 2 ab
92 ± 2 de
89 ± 2
Average
70 ± 3
94 ± 2
97 ± 1
88 ± 2
96 ± 1
96 ± 1
94 ± 2
95 ± 2
93 ± 2
94 ± 1
80 ± 4
97 ± 2
80 ± 2
70 ± 2
76 ± 3
85 ± 3
74 ± 3
74 ± 3
79 ± 3
85 ± 2
87 ± 2
24
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. acrticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
All species
24˚C 2007
88 ± 2 abc
98 ± 0 bc
100 ± 0 c
82 ± 3 abc
93 ± 3 bc
97 ± 3 bc
83 ± 1 abc
94 ± 2 bc
96 ± 4 bc
92 ± 6 bc
94 ± 4 bc
96 ± 2 bc
86 ± 4 abc
79 ± 2 abc
92 ± 2 bc
79 ± 4 ab
67 ± 6 a
92 ± 1 bc
77 ± 8 ab
89 ± 4 bc
89 ± 2
Figure 6: COD removal (%) and SO 4 concentration (mg/L) for each species and unplanted control at 4˚C and 24˚C. 4˚C bars are
averages of means of three replicates from two 4˚C incubations in 2007 and 2008. 24˚C bars are means of three replicates during the
24˚C 2007 incubation. Error bars represent one standard error.
SO4 (mg/L)
COD Removal (%)
50
70
80
90
0
100
40C
240C
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. articus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
2
4
6
8
10
12
14
25
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. articus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
60
40C
240C
26
Sulfate Concentration
Like COD removal, sulfate concentrations during 20-day wastewater batches
varied among plant species and temperatures (Figure 1). Some plant species’ effects on
SO 4 concentration varied strongly by temperature, while others’ did not (Table 3,
Table 5). Many more species showed temperature dependence for SO 4 than for COD
removal. For all species with seasonal differences, SO 4 concentrations were higher at
cooler temperatures (Figure 6).
At 24˚C, a rapid decrease in SO 4 concentration occurred in all columns within the
first day of the batch (Figure 1). By day 3 at 24˚C, SO 4 concentrations in all planted and
control columns were low, ranging from 0 to 1.2 mg/L. SO 4 remained low with little
change in concentrations throughout the 20-day batch. Concentrations did not differ
significantly among treatments at 24˚C in 2007 except for a small difference between J.
arcticus and the control (Table 6). Concentrations were significantly different at 24˚C in
2006; however, this was the result of low variability of near-zero values (0.0 – 0.4 mg/L).
During the 4˚C incubation, initial declines in SO 4 concentrations were not as rapid as
at 24˚C, and concentrations remained above 5 mg/L for several species (Figure 1). On
day 6, SO 4 concentrations ranged from 0.3 to 14.6 mg/L in the 4˚C 2007 batch and from
0.2 to 10.5 mg/L in the 4˚C 2008 batch (Table 5, Table 6). Time series for the unplanted
control, H. jubatum, C. canadensis, P. virgatum, P. arundinacea, L. cinereus, T. latifolia,
I. missouriensis, and P. vulgaris followed a pattern similar to that seen at 24˚C with SO 4
concentrations declining to near 0 mg/L by day 6 and little change throughout the
remainder of the batch. Day-6 SO 4 concentrations in columns with these plants were not
27
significantly different from the unplanted control. In contrast, 4˚C 2007 day-6 SO 4
concentrations ranged from 3.4 – 12.7 mg/L in columns planted with sedges, rushes, and
D. cespitosa; all of these differed significantly from the controls except C. microptera.
Table 5: SO 4 concentrations (mg/L) on day 6 of 4˚C batches in 2007 and 2008 and the
24˚C batch in 2007. SO 4 concentrations are means of three replicates ± one standard
error. Values with an asterisk (*) are significantly different from other values in the same
row (p<0.05, Repeated Measures ANOVA). Influent wastewater SO 4 concentrations
averaged 14 ± 0.5 mg/L.
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. acrticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
4 ˚C 2007
0.4 ± 0.1
7.7 ± 2.4
12.7 ± 0.6
3.4 ± 2.0
14.6 ± 1.2
12.1 ± 0.8
7.8 ± 1.0
10.1 ± 1.9
11.9 ± 2.6
0.6 ± 0.2
8.4 ± 1.7
0.9 ± 0.5
0.7 ± 0.3
0.7 ± 0.2
1.7 ± 0.7
0.3 ± 0.0
0.6 ± 0.3
1.4 ± 1.0
2.4 ± 1.1
*
*
*
24 ˚C 2007
0.4 ± 0.0
0.3 ± 0.1 *
0.4 ± 0.0 *
0.4 ± 0.0
0.4 ± 0.0 *
0.7 ± 0.3 *
0.4 ± 0.0 *
0.5 ± 0.1 *
1.2 ± 0.7
0.4 ± 0.0 *
0.4 ± 0.0
0.6 ± 0.1 *
0.3 ± 0.2
0.4 ± 0.0
0.5 ± 0.1
0.4 ± 0.0
0.4 ± 0.0 *
0.4 ± 0.0
0.4 ± 0.0
0.4 ± 0.0
4 ˚C 2008
0.2 ± 0.2
9.2 ± 1.9
8.4 ± 0.7
6.6 ± 2.2
7.9 ± 1.2
10.5 ± 2.1
6.8 ± 0.7 *
8.4 ± 0.3
6.4 ± 3.0
8.0 ± 2.2
1.3 ± 0.5
7.8 ± 0.9
2.6 ± 0.9
1.0 ± 0.5
0.4 ± 0.2
2.1 ± 0.7
5.4 ± 0.4 *
0.6 ± 0.3
1.3 ± 0.9
0.9 ± 0.2
Day-6 SO 4 concentrations in the columns containing sedges, rushes, D. cespitosa,
and P. arundinacea varied by seasons, while other treatments showed no to little seasonal
change (Figures 2-5). Negative correlations between temperature and SO 4 were found
for sedges, rushes, D. cespitosa, and P. arundinace (Table 3). Correlations were not
significant for other species.
Table 6: SO 4 concentrations (mg/L) on day 6 of batches at all temperatures. SO 4 values (mg/L) are means of three replicates ±
one standard error. Within each column, values in bold are significantly different than the control (p<0.05, planned contrasts).
Values in a column with the same letter are not significantly different from each other (p≥0.05, Tukey HSD).
16˚C 2006
1.3 ± 0.4 a
0.0 ± 0.0 a
0.1 ± 0.1 a
0.0 ± 0.0 a
0.0 ± 0.0 a
3.3 ± 1.6 ab
4.0 ± 2.5 ab
0.0 ± 0.0 a
1.8 ± 0.7 ab
5.9 ± 1.0 b
0.2 ± 0.2 a
2.4 ± 1.1 ab
0.0 ± 0.0 a
0.8 ± 0.0 a
1.1 ± 0.6 a
0.0 ± 0.0 a
0.0 ± 0.0 a
1.4 ± 0.1 a
0.7 ± 0.1 a
0.0 ± 0.0 a
1.1 ± 0.4
8˚C 2006
0.0 ± 0.0 a
2.1 ± 2.1 abcde
9.2 ± 1.0 def
2.3 ± 2.3 abcde
10.6 ± 0.7 ef
8.8 ± 1.7 cdef
12.1 ± 0.4 f
3.6 ± 2.0 abcdef
8.3 ± 3.1 abcdef
8.3 ± 3.0 abcdef
0.3 ± 0.1 abc
8.6 ± 1.8 bcdef
3.1 ± 1.0 abcde
0.5 ± 0.5 abc
0.2 ± 0.2 ab
0.4 ± 0.0 abc
0.4 ± 0.1 abc
1.0 ± 0.8 abcd
5.3 ± 2.6 abcdef
1.1 ± 0.8 abcd
4.5 ± 0.9
4˚C 2007
0.4 ± 0.1 a
7.7 ± 2.4 bcdef
12.7 ± 0.6 f
3.4 ± 2.0 abcde
14.6 ± 1.2 f
12.1 ± 0.8
7.8 ± 1.0
10.1 ± 1.9
11.9 ± 2.6
0.6 ± 0.2
8.4 ± 1.7
0.9 ± 0.5
0.7 ± 0.3
0.7 ± 0.2
1.7 ± 0.7
0.3 ± 0.0
0.6 ± 0.3
1.4 ± 1.0
2.4 ± 1.1
5.4 ± 1.2
f
cdef
ef
f
a
def
ab
a
a
abcd
a
a
abc
abcd
8˚C 2007
0.7 ± 0.0 a
5.3 ± 1.2 bcde
6.6 ± 1.1 cde
2.0 ± 0.6 ab
8.0 ± 0.3 de
5.9 ± 1.3 cde
9.1 ± 1.1 e
1.8 ± 0.7 ab
7.1 ± 1.4 de
4.8 ± 0.9 bcd
0.8 ± 0.2 a
2.8 ± 0.6 abc
0.7 ± 0.0 a
0.7 ± 0.0 a
0.6 ± 0.0 a
0.9 ± 0.2 a
0.8 ± 0.1 a
0.8 ± 0.1 a
0.9 ± 0.1 a
0.8 ± 0.2 a
3.2 ± 0.7
16˚C 2007
0.8 ± 0.0 a
3.4 ± 0.8 a
1.0 ± 0.1 a
0.9 ± 0.1 a
3.0 ± 1.8 a
1.0 ± 0.1 a
2.7 ± 1.5 a
0.9 ± 0.0 a
1.9 ± 1.0 a
0.8 ± 0.0 a
0.8 ± 0.0 a
1.6 ± 0.4 a
1.4 ± 0.6 a
0.8 ± 0.0 a
0.8 ± 0.0 a
0.8 ± 0.0 a
0.8 ± 0.0 a
0.8 ± 0.0 a
0.8 ± 0.0 a
0.8 ± 0.0 a
1.3 ± 0.2
28
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. acrticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
All species
24˚C 2006
0.0 ± 0.0 a
0.3 ± 0.0 bc
0.3 ± 0.0 c
0.3 ± 0.0 c
0.3 ± 0.0 c
0.3 ± 0.0 c
0.3 ± 0.0 bc
0.3 ± 0.0 c
0.3 ± 0.0 c
0.3 ± 0.0 c
0.3 ± 0.0 bc
0.4 ± 0.1 c
0.3 ± 0.0 c
0.0 ± 0.0 a
0.1 ± 0.1 ab
0.3 ± 0.0 bc
0.3 ± 0.0 bc
0.0 ± 0.0 a
0.0 ± 0.0 a
0.3 ± 0.0 c
0.2 ± 0.0
Table 6 continued: SO 4 concentrations (mg/L) on day 6 of batches at all temperatures. SO 4 values (mg/L) are means of three
replicates ± one standard error. Within each column, values in bold are significantly different than the control (p<0.05, planned
contrasts). Values in a column with the same letter are not significantly different from each other (p≥0.05, Tukey HSD). Averages
from the right-hand column are for batches 16˚C 2006 through 4˚C 2008: the first batch reflected start-up effects and is excluded.
Averages were not compared statistically.
24˚C 2007
0.4 ± 0.0
0.3 ± 0.1
0.4 ± 0.0
0.4 ± 0.0
0.4 ± 0.0
0.7 ± 0.3
0.4 ± 0.0
0.5 ± 0.1
1.2 ± 0.7
0.4 ± 0.0
0.4 ± 0.0
0.6 ± 0.1
0.3 ± 0.2
0.4 ± 0.0
0.5 ± 0.1
0.4 ± 0.0
0.4 ± 0.0
0.4 ± 0.0
0.4 ± 0.0
0.4 ± 0.0
0.5 ± 0.0
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
16˚C 2007
0.4 ± 0.1 a
6.0 ± 0.9 bcde
7.2 ± 1.1 cde
1.5 ± 1.0 ab
7.8 ± 0.6 cde
9.4 ± 0.9 e
4.5 ± 0.9 abcde
3.5 ± 1.1 abcd
8.2 ± 1.4 de
7.2 ± 2.0 cde
1.6 ± 1.4 ab
1.7 ± 0.2 ab
2.6 ± 1.6 abc
0.8 ± 0.4 ab
1.8 ± 0.8 ab
0.3 ± 0.1 a
0.9 ± 0.5 ab
0.4 ± 0.1 a
4.3 ± 0.8 abcde
3.1 ± 0.9 abcd
3.8 ± 0.7
8˚C 2007
0.6 ± 0.2 ab
4.8 ± 1.8 abcd
7.4 ± 0.9 abcd
4.3 ± 2.0 abcd
7.6 ± 1.8 abcd
10.6 ± 0.9 d
6.6 ± 2.0 abcd
8.7 ± 1.3 cd
7.9 ± 3.4 abcd
8.2 ± 0.9 bcd
0.6 ± 0.2 ab
2.1 ± 0.5 abc
0.6 ± 0.1 ab
0.5 ± 0.3 ab
0.2 ± 0.2 a
2.4 ± 0.7 abc
3.9 ± 0.8 abcd
0.3 ± 0.1 ab
3.4 ± 1.5 abcd
1.9 ± 1.2 abc
4.3 ± 0.8
4˚C 2008
0.2 ± 0.2 a
9.2 ± 1.9 bc
8.4 ± 0.7 abc
6.6 ± 2.2 abc
7.9 ± 1.2 abc
10.5 ± 2.1 c
6.8 ± 0.7 abc
8.4 ± 0.3 abc
6.4 ± 3.0 abc
8.0 ± 2.2 abc
1.3 ± 0.5 ab
7.8 ± 0.9 abc
2.6 ± 0.9 abc
1.0 ± 0.5 ab
0.4 ± 0.2 a
2.1 ± 0.7 abc
5.4 ± 0.4 abc
0.6 ± 0.3 abc
1.3 ± 0.9 ab
0.9 ± 0.2 abc
5.0 ± 0.8
Average
0.5 ± 0.1
4.3 ± 1.0
5.9 ± 1.4
2.4 ± 0.7
6.7 ± 1.6
6.3 ± 1.5
6.5 ± 1.4
3.9 ± 1.1
5.9 ± 1.1
6.2 ± 1.3
0.7 ± 0.1
4.0 ± 1.0
1.4 ± 0.4
0.7 ± 0.1
0.6 ± 0.2
1.0 ± 0.3
1.4 ± 0.6
0.7 ± 0.1
2.0 ± 0.6
1.3 ± 0.3
3.2 ± 0.6
29
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. acrticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia
All species
30
Relationships Between COD Removal and Sulfate Concentration
The species which showed the greatest and most consistent COD removal
throughout the year had the highest SO 4 concentrations at colder temperatures (Figures 25, Figure 6). Columns with less than 80% average COD removal had SO 4 concentrations
consistently near 0 mg/L. Relationships between COD removal and SO 4 concentration
were strongest at colder temperatures (Figure 7). The strongest relationship, with R2 =
0.85, occurred at 4˚C in both 2007 and 2008. The weakest relationship with R2 = 0.10,
occurred during the 24˚C 2007 period. The weak relationship was a result of the low
variability of SO 4 concentration among species at warm temperatures. The magnitude of
SO 4 variation and the strength of the relationship between COD removal and SO 4
concentration were intermediate at 8˚C and 16˚C.
SO 4 (mg/L)
10
8
4oC
12
24oC
y = 0.002e0.015x
p<0.0001
R2 = 0.85
10
y = 0.007x - 0.18
p=0.17
R2 = 0.1
SO 4 (mg/L)
12
6
4
2
8
6
4
2
0
0
60
70
80
90
COD removal (%)
100
60
70
80
90
100
COD removal (%)
Figure 7: Relationships between COD removal (%) and SO 4 concentrations (mg/L) at
4˚C and 24˚C. Each point represents day-6 measurements for one species or unplanted
controls. Values at 4˚C are averages of 4˚C 2007 and 2008 batches. Values at 24˚C are
for the 24˚C 2007 batch.
31
Redox Potential
Redox potential was also influenced by plant species and seasons. At 24˚C in
2006 and 2007, the redox potential in all planted and control columns dropped sharply
within the first day and remained in the range of -225 to -275 mV throughout the 20-day
wastewater batches (Figure 1). In contrast, greater variation occurred during the 4˚C
batches. Redox potential ranged from -236 to 401 mV on day 6 of the 4˚C batch in 2007
and from -260 mV to 445 mV on day 6 of the 4˚C batch in 2008. At 4˚C in 2007 and
2008 the control columns, and columns with C. microptera, C. canadensis, H. jubatum,
L. cinereus, P. arundinacea, P. australis, P. virgatum, P.vulgaris, and T. latifolia
followed a pattern similar to that seen at 24˚C. Columns planted with other species
behaved differently at 4˚C and 24˚C. At 4˚C columns with C. aquatilis, C. bebbii, C.
praegracilis, C. utriculata, S. acutus, J. arcticus, J. torreyi, and D. cespitosa either
declined less initially or increased throughout the batch from low initial levels (Figure 1
shows patterns for representative species).
Redox potential varied little among species at warm temperatures but varied
substantially during cold temperature batches (Figures 2-5). Day-6 redox potential
remained below -200 mV and was virtually unchanged across seasons in controls and
with the majority of plant species: C. aquatilis, C. microptera, C. canadensis, H.
jubatum, L. cinereus, P. virgatum, P. arundinacea, P. australis, I. missouriensis, P.
vulgarsis, and T. latifolia. Species that showed greater variation and rose above -200 mV
by the sixth day of cold temperature batches were C. bebbii, C. nebrrascensis, C.
praegracilis, C. utriculata, S. acutus, J. arcticus, J. torreyi, and D. cespitosa; Eh rose
32
above -200 mV by the ninth day in columns planted with C. aquatilis. Columns planted
with these species also had average COD removal greater than 90% and average SO 4
concentrations greater than 3 mg/L.
Discussion
Patterns of COD removal, sulfate concentration, and redox potential during 20day batch incubations demonstrated that wastewater treatment was affected by both
seasons and plants. Most plants affected seasonal patterns to some degree, but plant
effects were species specific. In unplanted controls, seasonal temperature variation had
little to no influence on SO 4 concentrations and redox potential, and COD removal
declined at lower temperatures.
Columns containing some plant species had water chemistry patterns similar to
unplanted controls, while other plant species behaved very differently from the controls.
Many species had consistently high COD removal throughout the year; columns with
these species had the highest average COD removal and had elevated SO 4 concentrations
and, sometimes, redox potentials at lower temperatures. Other species had intermediate
or inconsistent seasonal COD removal, SO 4 concentrations, and redox potential.
Lower COD removal in the unplanted controls during cold temperature
incubations was presumably caused by a decrease in biological activity at low
temperatures as conventionally assumed for wastewater treatment (USEPA 1993, Kadlec
and Knight 1996). Planted columns that followed patterns similar to the control were
likely also influenced by the effect of cold temperatures on biological activity. The
33
constant high COD removal values seen year-round with other species suggest that those
plants can somehow offset the negative effects of cold temperatures.
One possibility is that increased oxygen availability stimulates microbial
processes that would otherwise be negatively affected by cold temperatures. It is well
documented that many wetland plants release oxygen from their roots (Armstrong 1964,
Armstrong 1971b, Brix and Schierup 1990, Gries et al. 1990, Kludze et al. 1994, Kludze
and DeLaune 1994, Sorrell and Armstrong 1994, Laskov et al. 2006, Matsui and
Tsuchiya 2006, Tanaka et al. 2007). The ability of plants to transport oxygen from shoots
to roots efficiently and release oxygen from roots into the rhizosphere has the potential to
influence wastewater treatment processes. Because domestic wastewater has abundant
organic carbon, an electron donor, treatment is often limited by oxygen, the most
energetically favorable electron acceptor (Reed and Brown 1992, Kadlec and Reddy
2001). Leakage of oxygen from roots can create oxidized zones (Armstrong 1971b)
within the predominantly reduced environment of the bulk wastewater solution,
supporting more efficient aerobic treatment of COD and nitrogen (Brix 1987).
Seasonal variation in root oxygen loss could account for the observation that
differences in COD removal between planted and unplanted columns were found for
most species at lower temperatures but were infrequent at warmer temperatures. The
amount of root oxygen loss may vary across seasons due to variation in a plant’s oxygen
assimilation and respiration. Armstrong (1971b) and Howes and Teal (1994) found
increases in root oxygen loss from Oryza sativa (rice) and Spartina alterniflora (smooth
cordgrass), respectively, when temperature was lowered. Gries et al. (1990) and Moog
34
and Bruggemann (1998) measured both root respiration and oxygen release from reeds
and sedges. They found that with decreasing temperatures, root respiration rates
decreased while radial oxygen loss increased. The interaction between temperature and
root oxygen loss has been studied for only a small number of species. The SO 4
concentration and Eh results of this study suggest indirectly that elevated oxygen release
at lower temperatures may be fairly common among wetland plants.
SO 4 and Eh measurements in this experiment support the hypothesis that root
oxygen loss is a mechanism for plant species and seasonal effects on wastewater. At 4˚C
a strong relationship between COD removal and SO 4 concentration was found; there was
also a weaker relationship between COD removal and Eh. Species with low COD
removal also had low SO 4 concentration and redox potential; this may have resulted from
having low or no oxygen release from roots and, consequently, insufficient oxygen to
support sulfide oxidation, and raise Eh. Species with high COD removal also had
elevated SO 4 concentration and redox potential, which indicates elevated oxygen levels
that would support sulfide oxidation and enhance overall root-zone oxidation status. In
contrast, COD removal was not related to SO 4 concentrations or Eh at 24˚C because SO 4
and Eh were uniformly low in spite of variation in COD removal rates among species.
Past constructed wetland studies have also found differences in water chemistry
and microbial processes that suggest greater oxygen availability in winter than summer.
Kadlec and Reddy (2001) found a slight increase in BOD removal at lower temperatures,
consistent with the slight increase in COD removal at colder temperatures measured in
columns planted with C. nebrascensis and C. utriculata in this study. Griffin et al.
35
(1999) measured a lower Eh, increased ammonification and total N, and sulfate reduction
leading to sulfide formation in the summer; in the winter they found elevated Eh,
increased N removal, and no sulfate reduction.
Although differences were most pronounced in winter, the presence of plants,
especially select species, enhanced wastewater treatment throughout the year compared
to unplanted columns. Variations between plant species’ COD removal during colder
temperatures is likely driven by differing amounts of root oxygen loss between species.
Past studies have found that plant species differ in the amount of oxygen released from
roots (Bedford et al. 1991, Sorrell 1999, Chabbi et al. 2000, Visser et al. 2000b, Wiessner
et al. 2002). Differences in plants’ physiological and morphological traits adaptive to
their environment may influence the amount of root oxygen loss.
Columns planted with C. aquatilis, C. bebbii, C. praegracilis, C. utriculata, S.
acutus, J. arcticus, J. torreyi, and D. cespitosa had the greatest COD removal throughout
the year, with experiment average removals over 90%. These plant species, with the
exception of S. acutus, are not currently widely used in constructed wetlands and are not
well researched. T. latifolia, P. australis, and P. arundinaceae are widely used and well
researched, yet these species provided less improvement in treatment than many of the
species studied (average COD removal across temperatures = 84%, 74%, and 83%,
respectively). Because these three species are weedy in wetland ecosystems and have
unexceptional performance, other species may be preferable. In addition to the species
studied here, other members of the Cyperaceae and Juncaceae are likely candidates; all
of the species with >90% COD removal, except D. cespitosa, were in these two families.
36
All species with high COD removal were classified as Obligate Wetland or Facultative
Wetland species by the U.S. Fish and Wildlife Service (U.S. Fish and Wildlife Service
1988), but some species with the same Wetland Indicator Status did not perform as well.
Interestingly, C. microptera, the Carex species with the lowest Wetland Indicator Status
(Facultative), had the lowest average COD removal and SO 4 concentration of the Carex
species throughout the experiment (86% and 2 mg/L).
Results of this study suggest that choosing appropriate plants can increase
constructed wetland efficiency throughout the year and may be especially important in
cold climates. In spite of the normal tendency for microbial activity to decline during
cold temperatures, select plant species may create conditions for constant or even
increased COD removal at low temperatures. However, species must be matched to
wastewater goals. Plant species which promote aerobic processes will increase the
efficiency of wetlands designed for organic carbon and nutrient removal from domestic
wastewater (Burgoon et al. 1995, Cronk and Fennessy 2001, Nivala et al. 2007). On the
other hand, constructed wetlands designed for sulfate reduction and metal removal
require limiting oxygen to promote low redox conditions (Machemer and Wildeman
1992, Nelson et al. 2006, Stein et al. 2007), and plants that do not facilitate oxidation
would be most effective.
Results from this study that demonstrate pronounced differences between species
in the winter justify further investigation of plants’ seasonal effects on wastewater
treatment. A limited number of studies have compared plant species (Gersberg et al.
1986, Coleman et al. 2001, Allen et al. 2002, Fraser et al. 2004, Picard et al. 2005, Stein
37
and Hook 2005, Iamchaturapatr et al. 2006, Akratos and Tsihrintizis 2007, Yang et al.
2007) and even fewer have examined plant influences across all seasons (Allen et al.
2002, Picard et al. 2005, Stein and Hook 2005, Akratos and Tsihrintizis 2007, Yang et al.
2007). This study is unique in that it examines seasonal effects of a large number of
species for potential use in constructed wetlands including many regionally native plants
as well as commonly used species. Greenhouse-scale studies with controlled
environments will continue to be valuable for studying underlying biological processes.
However, because small-scale greenhouse studies may not predict performance in the
field due to scale differences and environmental and wastewater variation, field studies
are needed. This study and Allen et al. (2002) provide evidence that start-up effects
occur in model wetlands: COD removal was lower in the experiments’ initial batches.
This may be because plants were still growing into their containers and microbial
communities were developing. Both greenhouse and field studies will provide more
useful information if observed over two or more years. Additionally, rather than studying
only monocultures, further research should investigate wastewater treatment with mixed
plantings (Callaway and King 1996, Coleman et al. 2001, Picard et al. 2005). Additional
studies which make comparisons between a large number of species, both at small scales
and in the field, and studies of mixed planting will collectively contribute to plant
selection for improving constructed wetland function.
38
PLANT INFLUENCES ON WASTEWATER TREATMENT
Introduction
Constructed wetlands (CWs) are a natural alternative to conventional methods of
wastewater treatment (Kadlec and Knight 1996). Wastewater is treated by
physiochemical and biological interactions between wastewater, wetland media,
microorganisms, and plants. Although plants are a major component of wetlands,
relatively little is known about their effects on the processes responsible for wastewater
treatment. A greater understanding of plants’ influences would allow designers to select
appropriate plant species to increase the efficiency of the desired treatment, whether it is
organic carbon, nutrient, or heavy metal removal.
While comparisons of unvegetated and vegetated wetlands have found increased
treatment with plants present (Coleman et al. 2001, Hunter et al. 2001, Lin et al. 2002,
Picard et al. 2005, Iamchaturapatr et al. 2006, Yang et al. 2007) and species-specific
effects (Gersberg et al. 1986, Allen et al. 2002, Fraser et al. 2004, Stein and Hook 2005,
Akratos and Tsihrintizis 2007), the mechanisms underlying those effects are still the
subject of speculation. Plants could improve treatment through direct uptake of nutrients,
but plant nutrient content cannot account for the enhanced nutrient removal from
wastewater (Gersberg et al. 1986, Bachand and Horne 2000, Lin et al. 2002, Akratos and
Tsihrintizis 2007). Instead, influencing microbial activity is believed to be plants’
primary role in treatment wetlands by creating attachment sites and releasing exudates
and oxygen (Reed and Brown 1992, Brix 1997, Tanner 2001).
39
A plant’s influence on wastewater and microbial activity occurs primarily in the
rhizosphere, the area immediately surrounding the root that is influenced by root exudates
(Lynch and Whipps 1990). Abundant microbial communities, increased chemical
processing, and elevated dissolved oxygen have been measured in the root zones in
constructed wetlands (Reddy et al. 1989, Hatano et al. 1993, Bodelier et al. 1996, Ottova
et al. 1997, Weiss et al. 2003, Bezbaruah and Zhang 2004, Munch et al. 2005, Vacca et
al. 2005, Gagnon et al. 2006). These measurements provide evidence that roots increase
microbial processes due, in part, to oxygen release (Armstrong 1964, Wiessner et al.
2002, Bezbaruah and Zhang 2004). However, past studies measuring oxygen release
have conflicting results with roots either taking up or releasing oxygen (Armstrong
1971b, Brix and Schierup 1990, Gries et al. 1990, Bedford et al. 1991, Kludze et al. 1994,
Kludze and DeLaune 1994, Sorrell and Armstrong 1994, Sorrell 1999, Chabbi et al.
2000, Visser et al. 2000b, Laskov et al. 2006, Matsui and Tsuchiya 2006, Tanaka et al.
2007). There is disagreement regarding the potential influence of rhizodeposition – the
transfer of substances from roots to the surrounding medium – on the bulk soil or water
(Bedford et al. 1991, 1994; Sorrell and Armstrong 1994).
The debate over root influences in soil and water results partly from differences in
study methods and scale. Most studies of root effects involve short-term and small-scale
measurements of isolated roots, as opposed to studies that examine the net effect of the
whole plant in a wetland system taking into account the temporal and spatial variability
of roots acting as a source and a sink of oxygen (Bedford et al. 1991, 1994). While
isolated measurements in the rhizosphere provide evidence of the mechanisms underlying
40
plant influences on the bulk solution, the extent of these influences on wastewater
treatment also needs to be determined at appropriate scales.
Comparisons among COD removal, sulfate concentration, and redox potential
(Eh) suggest that plants do affect oxygen-dependent processes in constructed wetland
microcosms with mature plants (Chapter 2, Allen et al. 2002). COD removal from
wastewater declined at cold temperatures in microcosms planted with some plant species
but remained high at all temperatures with other species. SO 4 and Eh indicated that
conditions were highly reduced at all times in microcosms which had a decline in COD
removal at cold temperatures. In contrast, SO 4 and Eh indicated that a degree of
oxidation occurred in microcosms with consistent high COD removal. Allen et al.
(2002), Hook et al. (2003), and Stein and Hook (2005) hypothesized that the latter plant
species release greater amounts of oxygen in the winter when root respiration slows and
that this offsets the negative effects of temperature on COD breakdown by microbes.
If root oxygen loss does affect wastewater treatment in this way, plant species
could potentially be selected to increase the efficiency of COD removal and other
oxidative treatment processes based on the quantity of oxygen released from their roots.
Since root oxygen loss is problematic to measure and discrepancies occur with different
experimental methods (Sorrell and Armstrong 1994), other plant traits might prove to be
more useful for plant selection; possibilities include flooding tolerance, Wetland
Indicator Status (WIS), and root morphology, each of which is related to plants’ abilities
to avoid flooding stress caused by anoxia.
41
Plants differ in their physiological and morphological traits and ecological
distribution as a result of adaptation, or lack of adaptation, to wetland environments
(Laan et al. 1989, Sorrell et al. 2000, Visser et al. 2000a). Plants in wetland ecosystems
must survive in water-saturated substrates which provide inadequate oxygen for root
respiration and expose plants to reduced substances that can be toxic (Ponnamperuma
1972). Wetland plants have various traits and mechanisms for avoiding or tolerating
internal anoxia when flooded (McManmon and Crawford 1971, Drew 1997, Crawford
2003). In some species, oxygen is transported from shoots to roots, where it is used in
respiration; some of the oxygen may be released to the surrounding soil or water,
resulting in the oxidation of reduced, potentially harmful substances (Armstrong 1971b).
Thus, individual species adaptations to anoxic conditions may have important
implications for wastewater treatment processes affected by oxygen availability and may
also help explain species-specific effects on biogeochemical processes in wastewater
(Chapter 2).
My study investigated root oxygen loss (ROL) as a mechanism contributing to
plants’ influence on wastewater treatment in constructed wetlands and the potential of
ROL and related plant traits to guide species selection for use in CWs. Seasonal and
plant effects on water chemistry in wetland microcosms (“columns”) containing
monocultures of 19 different plant species were investigated in Chapter 2. In the study
presented here, root oxygen loss was measured while plants were dormant and while
plants were growing and compared to COD removal rates reported in Chapter 2.
Relationships with traits that may be related to ROL – aerenchyma, root diameter
42
distribution, flooding tolerance, Wetland Indicator Status, and family or genus – were
also investigated.
Materials and Methods
Species Selection
Nineteen plant species were selected so that species associated with different
hydrologic regimes and flooding tolerances could be compared. The Wetland Indicator
Status, a ranking that indicates the probability that a plant grows in a wetland, was used
to choose plants with varying degrees of adaptation to the wetland environment (U.S.
Fish and Wildlife Service 1988). Additionally, plant species were selected based on their
abundance in North America, particularly in the region surrounding Montana, their use
by professionals in the field of wetland restoration, and their use in past constructed
wetland research. The 19 species included sedges, rushes, grasses, and forbs (Table 7).
The plant species were either purchased from nurseries or transplanted from the field.
Plants were grown and experiments were conducted in the Plant Growth Center at
Montana State University in Bozeman, MT.
Root Oxygen Loss
Experimental Design: Root oxygen loss was measured on a subset of the 19
species during the winter and summer. Three replicates of an unplanted control and 14
plant species were measured in January 2007, when the greenhouse temperature was 4˚C
and plants were dormant (Table 7). Four replicates of the control and 10 plant species
43
were measured in July 2007, when the plants were actively growing and the greenhouse
temperature was 24˚C.
Table 7: Plant species studied and their Wetland Indicator Status. Obligate Wetland
species occur in wetlands >99% of the time, Facultative Wetland species 67-99% of the
time, and Facultative species 34-66% of the time. Root oxygen loss was measured at 4˚C
and/or 24˚C for species indicated with an X in the respective columns.
4˚C
24˚C
Family
Species
Common name
ROL
ROL
Obligate Wetland Species
Cyperaceae Carex aquatilis Wahlenb.
Carex bebbii Olney ex Fernald
Carex nebrascensis Dewey
Carex utriculata Boott
Typhaceae
Schoenoplectus acutus (Muhl.
ex Bigelow) A. Love & D. Love
Typha latifolia L.
Facultative Wetland Species
Cyperaceae Carex praegracilis W. Boott
Juncaceae
Poaceae
Iridaceae
Cyperaceae
Poaceae
Lamiaceae
Juncus arcticus Willd.
Juncus torreyi Coville
Calamagrostis canadensis
(Michx.) P. Beauv.
Deschampsia cespitosa (L.) P.
Beauv.
Phalaris arundinacea L.
Phragmites australis (Cav.)
Trin. ex Steud.
Iris missouriensis Nutt.
Facultative Species
Carex microptera Mack.
Hordeum jubatum L.
Leymus cinereus (Scribn. &
Merr.) A. Love
Panicum virgatum L.
Prunella vulgaris L.
water sedge
Bebb’s sedge
Nebraska sedge
Northwest
Territory sedge
hardstem bulrush
X
X
X
X
X
X
broadleaf cattail
X
X
clustered field
sedge
arctic rush
Torrey’s rush
bluejoint
X
X
X
X
X
X
X
tufted hairgrass
X
X
smallwing sedge
foxtail barley
basin wildrye
X
X
X
X
switchgrass
common selfheal
X
X
X
reed canarygrass
common reed
Rocky Mountain
iris
44
Before root oxygen loss was measured, plants were grown in saturated conditions.
Plants used for winter measurements were planted in May 2006 and grew until January
2007. Plants used for summer measurements were planted in April 2007 and grew until
July 2007. Nursery or field soil was removed from the roots, plants were trimmed for
uniform root and shoot biomass, and planted in sand (20/30 sand, Bozeman Brick Block
and Tile, Bozeman, MT) in Deepots ® (5.1 cm diameter x 22.9 cm, Stuewe & Sons, Inc.,
Corvallis, OR). Deepots ® were placed in Rubbermaid ® bins (51 cm long x 37 cm wide
x 31 cm high) and bins were filled with ¼ strength Hoagland’s nutrient solution so that
sand was saturated. Tap water was added to the base of the bins through a continuous
water delivery system to replace water losses from evapotranspiration. Every two weeks,
bins were drained and plants were randomized between and within bins to provide
uniform average conditions for plant growth. Bins were refilled with ¼ strength
Hoagland’s nutrient solution. Photoperiod followed seasonal light availability; no
supplemental lighting was used. The greenhouse temperature varied between 4, 8, 16,
and 24˚C in 60-day intervals to mimic seasonal cycles.
Root Oxygen Loss Measurements: Root oxygen loss was measured
colorimetrically by the titanium citrate method described by Kludze et al. (1994).
Titanium-citrate solution was prepared according to methods of Zehnder and Wuhrmann
(1976), using water which was deoxygenated by purging with nitrogen gas. Oxygen
contamination was avoided by creating solutions in an anaerobic glovebox (Protector
Controlled Atmosphere Glove Box, Labconco Corporation, Kansas City, MO). Ti3+
citrate solution was created by mixing 300 mL of 0.2 M sodium citrate solution, 30 mL of
45
1.16 M titanium chloride, and saturated sodium carbonate to adjust the pH to 5.6. This
solution was then diluted by a factor of ten with deoxygenated water.
In the greenhouse, sand was removed from plant roots, and an individual plant
and its sampling tube were suspended in a glass jar (Ball canning jars, 12 oz.). A known
amount of Ti3+ citrate solution was poured into the glass jar to immerse the roots.
Paraffin oil was applied at the root-shoot junction to hinder atmospheric oxygen
contamination of the root zone. Control treatments had the same apparatus with no plant.
After 24 hours, a 5 ml solution sample was extracted and stored in a Vacutainer
(BD Vacutainers, Franklin Lakes, NJ). Absorbance of the solution was measured at 527
nm in a spectrophotometer (Spectronic Genesys 5, Spectronic Instruments, Rochester,
NY). Oxygen concentrations were extrapolated from a standard curve (Oxygen =
-0.0011(Absorbance) + 0.3738) and the whole-plant oxygen loss was calculated using the
following equation:
ROL = v (p - c)
where ROL = root oxygen loss, µmol O 2 plant-1 day-1; v = standardized volume of Ti3+
citrate solution (volume of Ti3+ citrate solution for individual measurments / largest
volume of solution used in the experiment); p = oxygen concentration in the root solution
after 24 hours with plants, µmol O 2 plant-1; c = oxygen concentration in the root solution
after 24 hours in the control without plants, µmol O 2 plant-1. Roots and shoots were
dried to constant weight so that root oxygen loss measurements could be calculated per
gram of dry root.
46
Root Anatomy and Morphology
Experimental Design: Plant material was purchased or collected from the field in
April 2007. Plants were grown as described above until October 2007, at which time
they were harvested for root measurements.
Aerenchyma Measurement: The proportion of aerenchyma tissue was measured
in three roots from two replicates of C. canadensis, C. utriculata, D. cespitosa, S. acutus,
and T. latifolia following the methods of Visser and Bogemann (2003). Cross sections
were taken from 2-3 cm segments from the stem-root junction, stained with Cotton Blue,
and photographed with a Nikon Coolpix 990 and Eclipse E600 photomicroscope (Nikon
Inc., Japan). The diameters of the inner and outer root tissues and aerenchyma tissue
were measured using Adobe Photoshop CS3 (Adobe Systems Inc., San Jose, CA). The
proportion of aerenchyma tissue was calculated by dividing the area of aerenchyma tissue
by the area of the root cross section.
Root Measurement: Root morphology was measured for two replicates of five
plant species, C. canadensis, C. utriculata, D. cespitosa, H. jubatum, and S. acutus.
Dried roots from plants previously used to measure ROL were rehydrated and roots were
scanned using WinRhizo Pro Version 2002c and XLRhizo Version 2003a software
(Regent Instruments, Inc., Montréal, QE, Canada). Roots were divided into four diameter
classes ranging from 0.0 ≤ 0.60 mm up to >2.25 mm. The proportion of root length,
surface area, and volume in each diameter class was calculated.
47
Flooding Tolerance
To test the effects of flooding on plant growth, the 19 plant species were
subjected to two different hydrologic regimes. Four replicates of each species were
drained freely and four replicates were flooded to 5 cm above the surface of the soil.
Redox potential three centimeters below the soil surface averaged -30 mV in 12 flooded
pots and 245 mV in 12 drained pots.
Plant material was purchased or collected from the field, cut for uniform root and
shoot biomass, and planted in May 2006 in a 1:1:2 topsoil:peat:sand mixture in Treepots
® (11.4 cm x 11.4 cm x 35.6 cm, Stuewe & Sons, Inc., Corvallis, OR). Throughout the
experiment, the greenhouse temperature was maintained at 25˚C for 16 hours during the
day and 16˚C for 8 hours during the night. No nutrients or supplemental light were
provided. Plants were randomly assigned positions in the greenhouse and rotated to a
new position every two weeks to minimize differences in growing conditions. After five
months of growth, plants were harvested. Root and shoot tissues were separated,
cleaned, dried to constant mass, and weighed.
Statistical Analysis
Differences among plant species’ root oxygen loss, proportion of aerenchyma
tissue, and proportion of roots in each diameter class were determined by ANOVA (SPSS
Version 15.0, SPSS Inc., Chicago, IL). Planned contrasts were used to test whether
oxygen concentrations in planted treatments were different than in the unplanted controls.
Post hoc Tukey HSD comparisons were used to test for differences between plant species
in all experiments. Differences in shoot biomass, root biomass, and total biomass
48
between drained and flooded treatments were analyzed for each plant species
independently with a t-test using SPSS (SPSS, Inc., Chicago, IL Version 15.0). The
significance level was set at α = 0.05.
Pearson’s correlation (SPSS Version 15.0, SPSS Inc., Chicago, IL) was used to
test for associations between COD removal and root oxygen loss. COD removal data
from the 4˚C 2007 batch (Chapter 2) was correlated to the winter root oxygen loss
measurements while COD removal from the 24˚C 2007 batch was correlated to the
summer root oxygen loss measurements. To determine whether the correlation was
significant, the null hypothesis ρ = 0 was tested at a significance level of α = 0.05.
Least squares linear regression (SPSS, Inc., Chicago, IL Version 15.0) was used
to analyze relationships between COD removal (Chapter 2) and Wetland Indicator Status
and flooding tolerance for all species. COD removal data from the 4˚C 2007 incubation
(Chapter 2) was used. Wetland Indicator Status and flooding tolerance were treated as
categorical variables: facultative species were assigned a value of one, facultative
wetland species assigned a value of two, and obligate wetland species assigned a value of
three. Plants with greater biomass when flooded were assigned a value of one, and plants
with no biomass increase when flooded were assigned a value of two. Relationships
were considered significant with α = 0.05. The coefficient of determination (R2) was
obtained to determine the strength of the relationship.
49
Results
Root Oxygen Loss
During the winter, root oxygen release per plant and per unit mass ranged from
0.0 µmol O 2 plant-1 day-1 for the control to a maximum of 85.7 µmol O 2 plant-1 day-1 for
D. cespitosa and from 0.0 to 26.2 µmol O 2 g-1 dry root day-1 for the control and H.
jubatum, respectively (Table 8). Ten of 14 plant species released more oxygen than the
control based on whole plant measurements. Nine species released more oxygen than the
control when ROL was expressed on root mass basis. The plants identified as releasing
more oxygen differed between the two methods for calculating ROL because of the large
range in root biomass. Under the same growing conditions, root mass for individual T.
latifolia planted averaged 15.0 grams and H. jubatum roots averaged 0.77 grams.
During the summer, ROL ranged from -0.7 to 45.3 µmol O 2 plant-1 day-1 for the
control and S. acutus and from -0.7 to 25.6 µmol O 2 g-1 dry root day-1 for the contral and
S. acutus, respectively. Four of 10 species released more oxygen than the control based
on whole plant measurements, and two of 10 released more based on dry root
measurements. During the summer, there was less variation among plant species and
fewer species released more oxygen than the control. Summer and winter measurements
fell within the range of oxygen release measured in past studies: -270 µmol O 2 g-1 dry
mass h-1 (oxygen uptake) to 126 µmol O 2 g-1 dry mass h-1 (oxygen loss) (Bedford et al.
1991, Kludze et al. 1994, Sorrell 1999, Laskov et al. 2006).
So that these measurements could be compared to other studies, the amounts of
oxygen released by D. cespitosa and S. acutus were expressed based on root surface area
Table 8: Root oxygen loss at 4˚C and 24˚C. 4˚C ROL values (reported as µmol O 2 plant-1 day-1 and µmol O 2 g-1 dry root day-1) are
means of three replicates ± one standard error. 24˚C ROL values (reported as µmol O 2 plant-1 day-1 and µmol O 2 g-1 dry root day-1)
are means of four replicates ± one standard error. Values in a column with an asterisk (*) are significantly different than the control
(p<0.05, planned contrast).
4˚C
24˚C
Dry
root
(g)
5.7
4.8
4.1
3.8
3.0
2.4
5.6
15.0
1.5
0.8
1.9
1.6
1.9
1.1
Root oxygen loss
µmol O 2 plant-1 day-1
85.7 ± 5.5
73.4 ± 8.4
67.9 ± 7.0
60.3 ± 3.1
55.3 ± 17.7
47.7 ± 15.3
40.0 ± 23.3
35.2 ± 4.5
26.6 ± 5.1
25.5 ± 16.1
20.0 ± 14.2
17.7 ± 11.6
17.6 ± 0.5
13.1 ± 2.2
0.0 ± 1.5
*
*
*
*
*
*
*
*
*
*
Root oxygen loss
µmol O 2 g-1 dry root day-1
20.0 ± 5.7
17.6 ± 4.8
16.4 ± 1.3
16.2 ± 2.1
18.4 ± 5.5
23.2 ± 9.5
6.7 ± 2.4
3.0 ± 1.3
16.9 ± 2.1
26.2 ± 13.1
8.2 ± 4.5
9.4 ± 5.9
9.9 ± 2.2
13.1 ± 2.9
0.0 ± 1.5
*
*
*
*
*
*
*
*
*
Soln.
used
(mL)
250
255
250
250
Dry
root
(g)
1.7
4.3
1.5
2.3
250
Root oxygen loss
Root oxygen loss
13.8 ± 11.3
45.3 ± 12.4 *
4.0 ± 2.2
26.8 ± 9.7 *
5.9 ± 7.4
25.6 ± 13.5 *
2.5 ± 1.2
12.0 ± 3.2
1.2
4.2 ± 4.3
2.8 ± 3.4
340
250
250
250
5.6
1.3
1.3
1.6
30.7 ± 19.9 *
-0.2 ± 2.7
0.1 ± 7.4
20.8 ± 5.1 *
5.0 ± 2.9
-0.4 ± 2.0
0.1 ± 5.8
13.5 ± 3.1 *
250
250
1.0
10.5 ± 5.1
-0.7 ± 1.6
µmol O 2 plant-1 day-1
µmol O 2 g-1 dry root day-1
7.9 ± 4.2
-0.7 ± 1.6
50
D. cespitosa
S. acutus
C. microptera
J. torreyi
C. nebrascensis
C. bebbii
C. aquatilis
T. latifolia
C. canadensis
H. jubatum
C. utriculata
P. virgatum
J. arcticus
C. praegracilis
Control
Soln.
used
(mL)
217
205
217
217
223
217
212
217
233
208
217
217
213
207
154
51
as determined with WinRhizo Pro Version 2002c and XLRhizo Version 2003a software
(Regent Instruments, Inc., Montréal, QE, Canada) and planted area based on the area of
the mason jar used in the root oxygen loss measurements. D. cespitosa released 2.1 x
10-6 g O 2 cm-2 root surface area day-1 and 0.034 g O 2 m-2 planted area day-1. S. acutus
released 1.6 x 10-6 g O 2 cm-2 root surface area day-1 and 0.018 g O 2 m-2 planted area
day-1. These values also fall in the range reported in the literature (Armstrong 1971b,
Brix and Schierup 1990, Gries et al. 1990, Visser et al. 2000b, Matsui and Tsuchiya
2006, Tanaka et al. 2007).
Root Anatomy and Morphology
The percentage of aerenchyma tissue measured below the root-shoot junction
varied between species (Table 9, Figure 8). The proportion of aerenchyma tissue ranged
from 12% for C. canadensis to 59% for S. acutus.
Table 9: Percent aerenchyma below the root-shoot junction. Values are means of six
replicates ± one standard error. Values in a column with the same letter are not
significantly different (p≥0.05, Tukey HSD).
Calamagrostis canadensis
Deschampsia cespitosa
Typha latifolia
Carex utriculata
Schoenoplectus acutus
Aerenchyma Tissue (%)
11.8 ± 7.9 a
38.3 ± 3.7 b
55.2 ± 4.2 bc
56.9 ± 3.7 bc
58.7 ± 3.4 c
For all species, the majority of roots were less than 0.6 mm in diameter. As seen
in Table 10, plants differed in the relative proportion of roots less than 0.6 mm and within
the range from 0.6 mm to 1.5 mm. C. canadensis and H. jubatum had a greater
52
proportion of roots with the smallest diameter than D. cespitosa and S. acutus, which had
a greater proportion of 0.6 to 1.5 mm diameter roots.
(a)
(b)
Figure 8: Root cross sections of (a) C. canadensis and (b) D. cespitosa.
Photo (a) shows no aerenchyma tissue; aerenchyma structure is present in (b).
Table 10: Length per root diameter class (%). Values are the mean ± one standard error
of two replicates. Values in a column with the same letter are not significantly different
(p≥0.05, Tukey HSD).
C. canadensis
H. jubatum
C. utriculata
D. cespitosa
S. acutus
Length per root diameter class (%)
0.0 ≤ 0.60 mm
0.60 ≤ 1.5 mm 1.5 ≤ 2.25 mm
96.2 ± 1.1 a
3.4 ± 0.9 c
0.3 ± 0.1 a
94.7 ± 0.7 a
4.9 ± 0.7 bc
0.3 ± 0.0 a
88.3 ± 2.8 ab
11.0 ± 2.5 ab
0.4 ± 0.2 a
86.7 ± 0.5 b
12.6 ± 0.5 a
0.5 ± 0.0 a
86.6 ± 0.5 b
12.5 ± 0.3 a
0.6 ± 0.1 a
> 2.25 mm
0.2 ± 0.1 a
0.1 ± 0.0 a
0.3 ± 0.1 a
0.2 ± 0.0 a
0.4 ± 0.0 a
Flooding Tolerance
The plant species varied widely in their tolerance and growth response to
hydrologic regimes. Three replicates of L. cinereus and I. missouriensis, and one
replicate of P. vulgaris died in the flooded treatment. Two replicates of T. latifolia died
53
in the drained treatment. No significant difference in biomass between flooded and
drained treatments occurred with C. bebbii, J. arcticus, H. jubatum, D. cespitosa, C.
canadensis, P. virgatum, and P. vulgaris. The total biomass of C. aquatilis, C.
microptera, C. nebrascensis, C. praegracilis, C. utriculata, S. acutus, J. torreyi, P.
arundinacea, P. australis, and T. latifolia was greater in the flooded treatment; L.
cinereus and I. missouriensis’s total biomass was greater in the drained treatment (Figure
9). Some plant species allocated biomass differently in flooded than drained conditions.
*
flooded biomass / drained biomass
*
4
3
*
*
2
*
*
*
*
*
*
1
**
C.
C. aq
C. mi uat
n cr ili
C. ebr opt s
pr asc era
C. aeg ens
ut rac is
ric il
is
S
. a ulat
P.
ar J. cu a
un to tus
d r
P. ina rey
au ce i
a
T. stra e
la lis
C. tifo
J. be lia
b
H arc bii
.
t
D ju icu
C. . ce bat s
ca sp um
i
P. nad tosa
e
v n
P. irga sis
I. L. vulg tum
m ci a
iss ne ri
ou re s
rie us
ns
is
0
Figure 9: Biomass of plant species when flooded divided by biomass when drained.
Plant biomass values with an asterisk (*) differed significantly between flooded and
drained treatments (p<0.05).
54
C. praegracilis, J. torreyi,and P. australis’s total biomass was greater in flooded
conditions because their shoot biomass was significantly greater; however, their root
biomass was not greater with flooding (data not shown). Additionally, all plant species
except H. jubatum, P. virgatum, L. cinereus, I. missouriensis, and P. vulgaris developed
adventitious roots in the upper soil profile under flooded conditions. The formation of
adventitious roots is an adaptation to flooded conditions that increases the diffusion of
gases between shoots and roots (Visser et al. 1996). A possible factor contributing to I.
missouriensis’s greater biomass in the drained treatment was that the plants purchased
were approximately 5 cm tall and were therefore submerged in the flooded treatment.
Relationships Between Plant Traits and Plant Effects on Wastewater Treatment
ROL and COD Removal: COD removal on day 6 of the 4˚C 2007 and 24˚C 2007
wastewater batches were compared to root oxygen loss per plant taken during the same
temperature regime. COD removal and ROL were positively correlated at 4˚C (p = 0.001
and r = 0.67) but not at 24˚C (p = 0.74 and r = 0.10) (Figure 10).
The amount of oxygen required to facilitate plants’ influence on COD removal
was estimated and compared to measured root oxygen loss. During the 2007 batch at
4˚C, the difference in COD removal between the unplanted control and D. cespitosa was
201 mg/L after six days, equivalent to 1.05 mmols O 2 L-1 days-1. Root oxygen loss for D.
cespitosa during the winter was 86 µmol O 2 plant-1 day-1, which would account for only
8% of the required oxygen needed to explain the difference in COD removal between
planted and control columns. Similarly, differences in COD removal between
100
100
90
90
80
80
70
240C
r = 0.10
40C
r = 0.67
60
0
20
40
60
80
0
20
40
60
70
60
COD removal (%)
COD removal (%)
55
80
Root oxygen loss (µmol O2 plant-1 day-1)
Figure 10: Correlation between COD removal and root oxygen loss at 4˚C and 24˚C.
Each point in the 4˚C plot represents one species or unplanted control’s day-6 COD
removal (%) from the 4˚C 2007 batch incubation and ROL measured in winter. Each
point in the 24˚C plot represents one species or unplanted control’s day-6 COD removal
(%) from the 24˚C 2007 batch and ROL measured in summer.
the unplanted control and C. utriculata were equivalent to 0.98 mmols O 2 L-1 days-1. C.
utriculata released 20 µmol O 2 L-1 days-1, which was only 2% of the oxygen required to
facilitate COD removal.
Wetland Indicator Status, Flooding Tolerance, and COD Removal: A relationship
was found between WIS and COD removal (p = 0.001 and R2 = 0.35). All Facultative
species had relatively low COD removal with maximum removal of 83%. Facultative
wetland species had a large range from 67% to 99% COD removal, and all obligate
species had relatively high COD removal from 83% to 97%.
No relationship was found between species’ flooding tolerance and COD removal
(p = 0.19 and R2 = 0.10). The average COD removal for plants which had greater
biomass when flooded was 87%. In comparison, the average COD removal for plants
with no significant biomass difference between flooded and drained treatments was 81%.
56
Discussion
To survive in water-saturated soil conditions, wetland plants have adaptations to
meet their oxygen needs for root metabolism; some of that oxygen is released to the
surrounding soil or water (Beals 1917, Armstrong 1971a, Crawford 2003). Many authors
(Gersberg et al. 1986, Griffin et al. 1999, Kadlec and Reddy 2001, Allen et al. 2002,
Fraser et al. 2004) attribute the observed differences in wastewater treatment between
planted and unplanted wetlands to root oxygen loss. This study and a parallel study
(Chapter 2) compared root oxygen loss and constructed wetland water chemistry across
numerous plant species and between winter and summer, and also examined relationships
with other traits that may be related to root oxygen loss. My results provided some
support for the hypothesis that constructed wetland processes and chemistry are related to
root oxygen loss and that root oxygen loss contributes to greater effects of plants in
winter.
Comparisons between COD removal in model constructed wetlands, root oxygen
loss measurements, and plant attributes allow for correlations but not causal relationships.
The strength of the relationships between these measurements may be affected by the
complexity of these factors, the accuracy of measurements or rankings, or the differences
in experimental conditions and scale. The variables studied were measured at different
scales, under different conditions, and/or at different times. COD was measured in model
wetlands with mature plants, biofilms, and gravel substrate; root oxygen loss was
measured from small, young plants in oxygen-scavenging solution in a small jar; flooding
57
tolerance was measured with young plants in potting soil; aerenchyma was measured
from young isolated roots; and root dimensions were measured from the plants in the root
oxygen loss experiment. The Wetland Indicator Status is a ranking from professionals’
judgment of ecological distribution in the field. This study approach was chosen to look
for general patterns across many species and to identify practical, easily measured indices
to guide plant selection.
Relationship Between Wastewater COD Removal and ROL
The amount of root oxygen loss measured from the subset of species varied
widely, especially in winter. Ten of 14 plant’s root oxygen loss were significantly
different than the unplanted control in the winter versus 4 of 10 in the summer. Species’
rankings for amount of root oxygen loss also differed between seasons. Root oxygen loss
would be expected to vary among species based on differences in anatomy, morphology,
and physiology, and between seasons due to seasonal growth and dormancy cycles
(Luxmore et al. 1970, Howes and Teal 1994, Sorrell et al. 2000).
COD removal and ROL were correlated in the winter. Species with higher ROL
generally had higher COD removal. This correlation was not found in the summer.
These results parallel those found for COD, SO 4 , and redox potential in model
constructed wetlands reported in Chapter 2. In winter, differences in COD removal were
related to two indicators of root-zone oxygen availability: SO 4 , which requires oxygen
for sulfide oxidation, and Eh, a measurement of the oxidation-reduction potential. In
Chapter 2, nine of 19 species tested had consistently high COD removal year-round
(average >90%) and showed cold season increases in SO 4 ; all but one of these species
58
also had periods of elevated Eh at low temperatures. Six of these species had statistically
significant ROL compared to controls in the winter: D. cespitosa, S. acutus, J. torreyi, C.
nebrascensis, C. bebbii, and C. aquatilis. They also had six of the seven highest ROL
rates. However, three of those nine species with high COD removal, J. arcticus, C.
praegracilis, and C. utriculata, did not have significant ROL. The other 10 species tested
in Chapter 2 had lower and more variable COD removal (70-88%) and a consistently
anaerobic root-zone year-round. Winter ROL was measured in five of these species; four
(C. microptera, T. latifolia, C. canadensis, H. jubatum) had significant ROL compared to
the unplanted control, but only one (C. microptera) had above-average ROL. P.
virgatum did not have significant ROL. Together, the relationships between COD
removal and ROL, SO 4 , and Eh suggest that increased root-zone oxygen supply does
enhance COD removal in the winter.
Seasonal differences in the amount of ROL and relationships between ROL and
processes that consume oxygen in the root-zone may be explained by plants’ seasonal
growth and dormancy cycles. When temperatures are low and plants are dormant, roots
may act less as oxygen sinks due to a reduction in their respiration, and this can results in
increased oxygen loss (Armstrong 1971b, Greis et al. 1990, Howes and Teal 1994, Moog
1998, Moog and Bruggemann 1998). This could explain the increase in BOD and COD
removal with decreased temperatures that Kadlec and Reddy (2001) observed as well as
the constant COD removal measured with some plant species in Chapter 2 and previously
reported for C. utriculata and S. acutus by Allen at al. (2002). According to this
interpretation, certain species can increase oxygen availability in the root-zone at lower
59
temperatures, and this can in turn offset the negative effect of cold temperatures on
aerobic microbial processes (Allen et al. 2002, Hook et al. 2003, Stein and Hook 2005).
On the other hand, if plants release less oxygen during the summer due to higher
respiration, they would not influence the root-zone oxidation as much.
Although there was evidence that differences in ROL among species and seasons
might be related to differences in COD removal, it was inconclusive. The correlation
between ROL and COD removal was statistically significant in winter, but not
particularly strong (r = 0.67). Additionally, measured amounts of oxygen released from
plant roots could only account for a small portion of the additional COD removed in
columns planted with species that most enhanced removal.
There are several possible reasons for the discrepancy between ROL measured in
vitro and COD removal in the columns. The columns used to measure COD removal had
more root mass per volume of solution than the jars used to measure root oxygen loss: 5
times greater for D. cespitosa and 12 times greater for C. utriculata. These differences in
root mass and the volume of the solution help explain some of the discrepancies between
measurements. When measuring COD removal and root oxygen loss, more root mass
would affect the results obtained, since measurements were taken from the bulk solution.
Root oxygen release in vitro was diluted over a relatively larger volume of solution. The
rhizosphere generally extends 1 to 30 mm from the root (Munch et al. 2005, Edwards et
al. 2006). Therefore, a greater proportion of the wastewater in the water quality
experiment was influenced directly by the roots. It is likely that the comparison between
root masses underestimates the differences between the rhizosphere influences on COD
60
removal and root oxygen loss. Greater root mass and the rhizosphere effect would
increase root oxygen loss values and could explain the COD removal difference between
planted and control columns.
It is not surprising that relationships between ROL and wastewater data are
ambiguous. The use of ROL measurements to explain processes at larger scales has been
questioned because of differences in media, plant size and age, and presence or absence
of organic matter and biofilms (Bedford et al. 1991 and 1994, Sorrell and Armstrong
1994). Relating wetland processes to ROL also depends on how ROL is characterized
(per plant, per unit of root mass, or per surface area). Although ROL measurements in
this study were consistent with other reported values (Armstrong 1971b, Brix and
Schierup 1990, Gries et al. 1990, Bedford et al. 1991, Kludze et al. 1994, Sorrell 1999,
Visser et al. 2000b, Matsui and Tsuchiya 2006, Laskov et al. 2006, Tanaka et al. 2007),
those values would be subject to the same questions about measurement methods and
scale. Discrepancies could also reflect the complexity of interactions between plants and
root-zone processes. For example, rhizodeposition and enhanced surface area for growth
of attached microbes may contribute to plant influences.
Relationships Between Wastewater COD Removal and Plant Attributes
Because in vitro root oxygen loss is difficult to measure, I investigated other
variables that could be related to ROL. These included: anatomical and morphological
traits that can affect oxygen transport directly; flooding tolerance and ecological
association with wetlands, which typically involve adaptations to maintain adequate
61
internal oxygen levels; and membership in plant families with a disproportionate number
of wetland species.
Aerenchyma tissue, which can help transport oxygen to roots and increase the
release of oxygen, and anatomical barriers to oxygen loss, such as suberized cells, can
differ among wetland species (Armstrong 1979, Jackson and Drew 1984, Justin and
Armstrong 1987, Armstrong and Armstrong 1988, Moog 1998, Jackson and Armstrong
1999, Chabbi et al. 2000, Visser et al. 2000b). Of the five species examined, C.
canadensis, the species with the lowest proportion of aerenchyma below the root-shoot
junction (12%) had the lowest COD removal rates and SO 4 concentrations in wastewater
incubations. The other four species had 38% to 59% aerenchyma; their COD removal
rates were intermediate to high and unrelated to the amount of aerenchyma.
Of five species characterized, those with proportionally more 0.6 - 1.5 mm
diameter roots (D. cespitosa, C. utriculata, and S. acutus) had >90% average COD
removal, while those with more 0.0 - 0.6 mm diameter roots (C. canadensis and H.
jubatum) averaged approximately 80% COD removal. Thicker roots had more
aerenchyma tissue, which would facilitate efficient gas transport. Visser et al. (2000)
found that the diameter of flooded roots increased as a result of increased aerenchyma.
Wetland Indicator Status, a rating indicating the occurrence of a plant species in a
wetland, is easily obtained from U.S. Fish and Wildlife Service lists. Because plants’
distributions and flood tolerances are influenced by differences in root anatomy,
morphology, and gas flow (Justin and Armstrong 1987, Laan et al. 1989, Sorrell et al.
2000), WIS may give some indication of a species’ capacity to transport oxygen to roots
62
and, potentially, into the surrounding medium. Comparing plants’ WIS and COD
removal revealed that Facultative species had relatively low but inconsistent COD
removal (70%-88%), Facultative Wetland species varied widely (74%-97%), and
Obligate Wetland species consistently had high COD removal (≥85%). Five out of six
Obligate Wetland species, three of eight Facultative Wetland species, and none of the
five Facultative species had ≥90% average COD removal.
Greenhouse ratings of flood tolerance showed a limited relationship to COD
removal. The two species that had significantly less biomass when flooded had the
lowest average COD removal. Species that had no significant biomass differences
between flooded or drained conditions and species with greater biomass in the flooded
condition had equally wide ranges of COD removal, approximately 75% to nearly 100%.
Four of seven species which had no significant biomass differences between flooded or
drained conditions had greater COD removal than the control, three had an average COD
removal >90%. Nine of ten species with greater biomass in the flooded condition had
greater COD removal than the control; six of them had an average COD removal >90%.
Additional observations made during these studies revealed differences between
plant species. Three of the selected 19 plant species, L. cinereus, P. vulgaris, and P.
viragatum would not be good candidates for constructed wetlands, since some replicates
of these species did not survive in the flooding-tolerance and wastewater treatment
experiments; these three species were among the species with the lowest average COD
removal. Photosynthesis was not quantified in this study during the winter; however,
photosynthesis has been measured during low temperature winters in coniferous trees
63
(Freeland 1944, Parker 1953). Since photosynthetic oxygen is transported to the
belowground roots, plant species that conduct photosynthesis during the winter possibly
assimilate more oxygen and could influence root oxygen loss. When the greenhouse
temperature was set at 4˚C, C. aquatilis, C. bebbii, C. nebrascensis, C. microptera, C.
praegracilis, C. utriculata, S. acutus, J. arcticus, J. torreyi, D. cespitosa, and P. vulgaris
had green leaves, the characteristic pigment of chlorophyll in photosynthetic organisms;
other species were brown. Species with green leaves, with the exception of P. vulgaris,
had the greatest average COD removal. While harvesting the flooding-tolerance
experiment, I made observations as to whether plants grown in flooded conditions grew
adventitious roots. Some species have the ability to grow new adventitious roots near the
soil surface for increased oxygen assimilation (Laan et al. 1989, Visser et al. 2000b,
Kyambadde et al. 2004). All species grew adventitious roots except for H. jubatum, P.
virgatum, L. cinereus, I. missouriensis, and P. vulgaris; these species without
adventitious roots were among the species with lowest average COD removal. Rooting
depth was observed while destructively harvesting three of 19 plant species at the end of
the water quality experiment. L. cinereus, which had the lowest average COD removal
among plant species, had root growth only in the top 10 cm of the microcosm. In
contrast, C. utriculata and D. cespitosa, which are two species with average COD
removal >90% grew roots to the bottom of the microcosm.
Plant family was related to COD removal as strongly as any other attribute. Most
species removed significantly more COD than unplanted controls at least some of the
time, but plants with the greatest influence on COD, SO 4 , and Eh were mostly members
64
of the Cyperaceae and Juncaceae. Plants with >90% average COD removal and
seasonally elevated SO 4 and Eh included hardstem bulrush and five of six sedges
(Cyperaceae), both rushes (Juncaceae), and one of seven grasses, D. cespitosa
(Poaceae). Plants with lower influence on COD removal and rootzone oxidation
included six grasses and one sedge, C. microptera; C. microptera was the only sedge
rated as Facultative. With one species each, no generalizations can be made for the
Iridaceae, Lamiaceae, and Typhaceae.
Selecting Appropriate Species for Constructed Wetlands
Species affect seasonal performance, with respect to COD removal, in cold
climates (Chapter 2) and this may be related to effects on root-zone oxygen. Chapter 2
results support the proposition that species selection is probably more important in cold
regions than warm. This study (Chapter 3) found large variations in ROL and plant
attributes related to ROL; these were related in differing degrees to differences in COD
removal, SO 4 , and Eh reported in Chapter 2. Although not definitive, these results
generally support the hypothesis that root oxygen loss is a mechanism that contributes to
plant species effects on water chemistry. They also suggest that plants’ adaptive traits for
living in a wetland environment influence their effects on wastewater treatment.
Although intuitive, this is not inevitable; for example, plant effects could reflect their
organic carbon production, support of biofilms, or other factors unrelated to oxygen.
Results from this study suggest that plant selection can help optimize wastewater
treatment. The appropriate species for a given treatment wetland depends on the limiting
factors and the chemical transformations needed for the relevant treatment processes.
65
Plant species that promote aerobic processes will increase the efficiency of wetlands
designed to remove organic compounds and nutrients from domestic wastewater
(Burgoon et al. 1995, Cottingham et al. 1999, Nivala et al. 2007). In cold climates, root
oxygen loss can apparently offset the expected effect of low temperatures on microbial
activity and maintain constant COD removal over the seasons. On the other hand,
constructed wetlands designed for sulfate reduction and metal removal require the
exclusion of oxygen to promote low redox conditions (Machemer and Wildeman 1992,
Nelson et al. 2006, Stein et al. 2007). Plants that release significant amounts of oxygen
should be avoided in this case.
Results provide insights into the mechanisms responsible for plants’ effects on
biogeochemical processes. Previous studies have shown that differences in adaptive
traits for living in the wetland environment can affect transport of oxygen to the rootzone (Justin and Armstrong 1987, Laan et al. 1989, Sorrell et al. 2000). Results reported
here and in Chapter 2 indicate that such differences can be important to plants’ effects on
biogeochemical processes that use oxygen. Few studies have examined variation of these
effects over enough species to draw generalizations. Basic understanding of how plants
enhance aerobic processes in wetlands could be advanced using a combination of field
studies and rhizosphere-scale sampling in wetland microcosms. Most of the species with
the highest COD removal rates in Chapter 2 were natives that are not widely used or well
tested. This study provides support for using native species and guidance for selecting
them. The influence of a diverse selection of species could further enhance constructed
66
wetland performance through differences in phenology, rooting patterns, and their
influence on root-zone processes.
The measurements of plant attributes used here to evaluate relationships with
COD removal differ in their ease of obtainability. Root oxygen loss measurements are
technically difficult and values in the literature are not comparable between plant species
due to different experimental designs (Sorrell and Armstrong 1994). Root characteristics
were also technically difficult to measure and values in literature could also be variable
depending on experimental conditions. Flooding tolerance measurements were time
consuming and not predictive. From a constructed wetland designer’s perspective,
Wetland Indicator Status and taxonomy would be the most easily obtained plant attributes
for selecting species. All Obligate Wetland species and members of the Cyperaceae and
Juncaceae families had greater COD removal than the control and comprised a large
proportion of species with COD removal ≥90%. These are likely plant groups that would
be appropriate for constructed wetlands but exceptions will certainly emerge. These
groups also include plants with intermediate COD removal: T. latifolia, an Obligate
Wetland species, had average COD removal of 85%, and C. microptera, a member of the
Cyperaceae but a Facultative species, had an average COD removal of 88%. D.
cespitosa, a Facultative Wetland species and a member of Poaceae, had the highest
average COD removal (97%) but is not in the mentioned plant groups. In spite of these
inconsistencies, family and WIS can still serve as “coarse filters” to identify candidates
for further evaluation.
67
CONCLUSION
Seasonal Plant Effects on Wastewater Treatment
Because plants influence the biology and chemistry around their roots in a zone
referred to as the rhizosphere, it has long been thought that they could influence
wastewater treatment in constructed wetlands. However, plants’ effects on wastewater
treatment have been debated due to contradictory results comparing treatment efficacy
between planted and unplanted wetlands.
My research showed that plants had seasonal effects on COD removal, SO 4
concentrations, and Eh, and that those effects were species-specific. COD removal in the
control declined in the winter, probably due to the decline of microbial activity at colder
temperatures. Cold temperature did not affect COD removal negatively in most planted
columns, except for those planted with I. missouriensis, L. cinereus, and P. virgatum.
COD removal remained constant with many species or, in the case of columns planted
with C. nebrascensis and C. utriculata, increased at low temperatures. These results
provide experimental support for the feasibility of constructed wetlands in cold climates,
adding to the evidence from case studies presented in Mander and Jenssen (2003) and
elsewhere.
The strong differences found among species suggest that specific plant species
could be used to improve wastewater treatment. Columns with C. aquatilis, C. bebbii, C.
nebrascensis, C. praegracilis, C. utriculata, S. acutus, J. arcticus, J. torreyi, and D.
cespitosa had high COD removal across all seasons, with an average over the entire
68
experiment greater than 90%. In contrast, columns planted with some species had lower
COD removal at 4˚C than at 24˚C, and in other columns, COD removal was mediocre
throughout the seasons with no apparent temperature influence. The year-round high
COD removal by select species suggests that these plants can offset the effects of cold
temperature that were observed in the control columns and that these plants have traits
that result in greater influence on wastewater treatment than other plants.
Positive correlations between plants’ COD removal and both SO 4 concentration
and Eh during colder temperatures indicated, indirectly, that species with high COD
removal have elevated root zone oxygen availability in the winter. Sustained aerobic
respiration by microbes may offset the negative temperature effects on microbial
processes and keep COD removal constantly high in columns planted with species that
enhance oxygen supply. Root oxygen loss could be the reason for both elevated oxygen
levels in select columns and the higher COD removal.
In contrast, in unplanted controls
or columns planted with other species for which COD removal declined, SO 4
concentrations and redox potentials remained low during cold temperatures.
Results from Chapter 3 showed that at 4˚C root oxygen loss per plant was
positively correlated with COD removal; species with greater root oxygen loss had
greater COD removal. No correlation existed at 24˚C. However, comparisons between
root oxygen loss values and the oxygen needed to account for elevated COD removal in
planted columns showed that root oxygen loss only accounted for a small fraction of the
oxygen required. Discrepancies may result from differences in root mass densities
between root oxygen loss and wastewater treatment experiments or from flaws in
69
measuring root oxygen loss. The reliability and meaning of root oxygen loss
measurements is debated (Bedford et al. 1991, Sorrel and Armstrong 1994) and
experimental methods can yield inaccurate oxygen measurements. Regardless, results
suggest that differences in root oxygen loss may be a factor in COD removal differences
between seasons and plant species. Since root oxygen loss measurements do not fully
explain patterns of COD removal, other mechanisms for plant and seasonal influences on
wastewater treatment should be investigated.
Attributes related to root oxygen loss were consistent with patterns of COD
removal. Root anatomy and morphology can influence root oxygen loss and, therefore,
could influence COD removal. Greater amounts of aerenchyma tissue and larger roots
are factors which would increase root oxygen loss (Justin and Armstrong 1987, Visser et
al. 2000b). Of the five species studied, C. canadensis had the least amount of
aerenchyma tissue and the greatest portion of small diameter roots, and the lowest COD
removal. The others had more aerenchyma tissue, larger diameter roots, and higher COD
removal.
Plant Selection for Constructed Wetlands
The presence of plants can improve wastewater treatment, and may be especially
important in cold climates. The average COD removal in unplanted model wetlands
across all seasons was 67% and COD removal declined during colder temperatures. In
contrast, the average COD removal in planted model wetlands was 87%. Unlike the
unplanted model wetlands, cold temperature generally did not have a negative effect on
70
COD removal in planted wetlands. COD removal remained relatively constant across
seasons in planted wetlands indicating that plants can promote more predictable
treatment.
Based on my results, species selection may also be used to increase the efficacy of
treatment. C. aquatilis, C. bebbii, C. praegracilis, C. utriculata, S. acutus, J. articus, J.
torreyi, and D. cespitosa had high COD removal throughout the year, with average COD
removals over 90%. These plant species, with the exception of S. acutus, are not
frequently used in constructed wetlands, but they had greater removal than three widely
used plants species, T. latifolia, P. australis, and P. arundinaceae, which had average
COD removal of 84%, 74%, and 83% COD removal, respectively. In addition T.
latifolia, P. australis, and P. arundinaceae are weedy in many North American wetland
ecosystems; other species may provide better wastewater treatment while enhancing
ancillary benefits including better habitat and a more aesthetic greenspace with a
diversity of species.
Comparing 19 plant species and unplanted model wetlands demonstrated that
certain groups of plants had similar patterns of COD removal (Table 11). Taxonomic
group and Wetland Indicator Status, which are easily obtained, can be useful for selecting
plant species that promote COD removal. All Cyperaceae and Juncaceae species used in
this study had greater COD removal than the unplanted wetland, generally better than
90%. Other members of these families that occur primarily in wetlands would also likely
be good candidates for treatment wetlands. All the species classified as Obligate Wetland
71
species (U.S. Fish and Wildlife Service 1988) had greater COD removal than the
unplanted control, again usually over 90%.
Future research could provide additional insight for selecting appropriate plant
species for the goals of the treatment wetland. Tests in large-scale wetlands in the field
would provide more realistic performance evaluations. Additionally, wastewater
treatment with mixed species plantings should be investigated (Callaway and King 1996,
Coleman 2001, Picard et al. 2005). Collectively, such studies will contribute to rational
plant selection for improving constructed wetland function.
Table 11: Summary of plant species studied and their attributes and influences on wastewater. Obligate Wetland species
occur in wetlands >99% of the time, Facultative Wetland species 67-99% of the time, and Facultative species 34-66% of the
time. COD removal averages (%) are for batches 16˚C 2006 through 4˚C 2008. SO 4 concentrations (mg/L) and Eh (mV)
increased in the winter and biomass (g) increased in flooded conditions for species indicated with an X in the respective
columns. Winter ROL values (µmolO 2 plant-1day-1) are averages of three replicates. ROL and aerenchyma (%) were measured
for only a subset of species.
Family
Cyperaceae
Cyperaceae
Cyperaceae
Cyperaceae
Cyperaceae
Cyperaceae
Cyperaceae
Juncaceae
Juncaceae
Poaceae
Poaceae
Poaceae
Poaceae
Poaceae
Poaceae
Poaceae
Iridaceae
Lamiaceae
Typhaceae
Obligate Wetland
Obligate Wetland
Facultative
Obligate Wetland
Facultative Wetland
Obligate Wetland
Obligate Wetland
Facultative Wetland
Facultative Wetland
Facultative Wetland
Facultative Wetland
Facultative
Facultative
Facultative
Facultative Wetland
Facultative Wetland
Facultative Wetland
Facultative
Obligate Wetland
Elevated
SO 4 in
winter
X
X
X
X
X
X
X
X
X
X
Elevated
Eh in
winter
X
X
X
X
X
X
Increased
biomass in
flooded
X
X
X
X
X
X
X
X
X
Winter ROL
µmolO 2
plant-1day-1
0.0
40.0
47.7
67.9
55.3
13.1
20.0
73.4
17.6
60.3
26.6
85.7
25.5
Aerenchyma
(%)
56.9
58.7
11.8
38.3
17.7
X
X
X
X
35.2
55.2
72
Species
Control
C. aquatilis
C. bebbii
C. microptera
C. nebrascensis
C. praegracilis
C. utriculata
S. acutus
J. arcticus
J. torreyi
C. canadensis
D. cespitosa
H. jubatum
L. cinereus
P. virgatum
P. arundinacea
P. australis
I. missouriensis
P. vulgaris
T. latifolia.
Wetland
Indicator Status
COD
removal
average ±
SE (%)
70 ± 3
94 ± 2
97 ± 1
88 ± 2
96 ± 1
96 ± 1
94 ± 2
95 ± 2
93 ± 2
94 ± 1
80 ± 4
97 ± 2
80 ± 2
70 ± 2
76 ± 3
85 ± 3
74 ± 3
74 ± 3
79 ± 3
85 ± 2
73
References
Akratos, C. S., and V. A. Tsihrintzis. 2007. Effect of temperature, HRT, vegetation and
porous media on removal efficiency of pilot-scale horizontal subsurface flow
constructed wetlands. Ecological Engineering 29:173-191.
Allen, W. C., P. B. Hook, O. R. Stein, and J. A. Beiderman. 2002. Temperature and
wetland plant species effects on wastewater treatment and root zone oxidation.
Journal of Environmental Quality 31:1010-1016.
Armstrong, W. 1964. Oxygen diffusion from the roots of some British bog plants. Nature
204:801-802.
Armstrong, W. 1971a. Oxygen diffusion from roots of rice grown under non-waterlogged
conditions. Physiologia Plantarum 24:242-247.
Armstrong, W. 1971b. Radial oxygen losses from intact rice roots as affected by distance
from apex, respiration and waterlogging. Physiologia Plantarum 25:192-197.
Armstrong, W. 1979. Aeration in higher plants. Advances in Botanical Research 7:225332.
Armstrong, J., and W. Armstrong. 1988. Phragmites australis - A preliminary study of
soil-oxidizing sites and internal gas-transport pathways. New Phytologist
108:373-382.
Bachand, P. A. M., and A. J. Horne. 2000. Denitrification in constructed free-water
surface wetlands: II. Effects of vegetation and temperature. Ecological
Engineering 14:17-32.
Barko, J. W., and R. M. Smart. 1983. Effects of organic-matter additions to sediment on
the growth of aquatic plants. Journal of Ecology 71:161-175.
Beals, C. C. 1917. The effect of aeration on the roots of Zea mays. Pages 177-180 In
Bennett, L.F. editor. Indiana Academy of Social Sciences, Indianapolis, IN.
Bedford, B. L., and D. R. Bouldin. 1994. On the difficulties of measuring oxygen release
by root systems of wetland plants - response. Journal of Ecology 82:185-186.
Bedford, B. L., D. R. Bouldin, and B. D. Beliveau. 1991. Net oxygen and carbon-dioxide
balances in solutions bathing roots of wetland plants. Journal of Ecology 79:943959.
Bezbaruah, A. N., and T. C. Zhang. 2004. pH, redox, and oxygen microprofiles in
rhizosphere of bulrush (Scirpus validus) in a constructed wetland treating
municipal wastewater. Biotechnology and Bioengineering 88:60-70.
74
Bodelier, P. L. E., J. A. Libochant, C. W. P. M. Blom, and H. J. Laanbroek. 1996.
Dynamics of nitrification and denitrification in root-oxygenated sediments and
adaptation of ammonia-oxidizing bacteria to low-oxygen or anoxic habitats.
Applied and Environmental Microbiology 62:4100-4107.
Brix, H. 1987. Treatment of waste-water in the rhizosphere of wetland plants - the rootzone method. Water Science and Technology 19:107-118.
Brix, H. 1997. Do macrophytes play a role in constructed treatment wetlands? Water
Science and Technology 35:11-17.
Brix, H. and H.H. Schierup. 1990. Soil oxygenation in constructed reed beds: The role of
macrophyte and soil-atmosphere interface oxygen transport. Pages 53-66 In
International Conference on the Use of Constructed Wetlands in Water Pollution
Control. Pergamon Press. Oxford, UK
Burgoon, P. S., K. R. Reddy, and T. A. Debusk. 1995. Performance of subsurface flow
wetlands with batch-load and continuous-flow conditions. Water Environment
Research 67:855-862.
Callaway, R. M., and L. King. 1996. Temperature-driven variation in substrate
oxygenation and the balance of competition and facilitation. Ecology 77: 11891195.
Chabbi, A., K. L. Mckee, and I. A. Mendelssohn. 2000. Fate of oxygen losses from
Typha domingensis (Typhaceae) and Cladium jamaicense (Cyperaceae) and
consequences for root metabolism. American Journal of Botany 87:1081-1090.
Coleman, J., K. Hench, K. Garbutt, A. Sexstone, G. Bissonnette, and J. Skousen. 2001.
Treatment of domestic wastewater by three plant species in constructed wetlands.
Water Air and Soil Pollution 128:283-295.
Cottingham, P. D., T. H. Davies, and B. T. Hart. 1999. Aeration to promote nitrification
in constructed wetlands. Environmental Technology 20:69-75.
Crawford, R. M. M. 2003. Seasonal differences in plant responses to flooding and anoxia.
Canadian Journal of Botany 81:1224-1246.
Cronk, J. K., and M. S. Fennessy. 2001. Wetland Plants: Biology and Ecology. Lewis
Publishers, London.
Drew, M. C. 1997. Oxygen deficiency and root metabolism: Injury and acclimation under
hypoxia and anoxia. Annual Review of Plant Physiology and Plant Molecular
Biology 48:223-250.
75
Edwards, K. R., H. Cizkova, K. Zemanova, and H. Santruckova. 2006. Plant growth and
microbial processes in a constructed wetland planted with Phalaris arundinacea.
Ecological Engineering 27:153-165.
Ernst, W. H. O. 1990. Ecophysiology of plants in waterlogged and flooded environments.
Aquatic Botany 38:73-90.
Faulkner, S. P., W. H. Patrick, and R. P. Gambrell. 1989. Field techniques for measuring
wetland soil parameters. Soil Science Society of America Journal 53:883-890.
Federal Register. 1980. "40 CFR Part 230: Section 404(b)(1) Guidelines for Specification
of Disposal Sites for Dredged or Fill Material," Vol 45, No. 249, pp 85352-85353,
US Government Printing Office, Washington, D.C.
Federal Register. 1982. "Title 33: Navigation and Navigable Waters; Chapter II,
Regulatory Programs of the Corps of Engineers," Vol 47, No. 138, p 31810, US
Government Printing Office, Washington, D.C.
Fraser, L. H., S. M. Carty, and D. Steer. 2004. A test of four plant species to reduce total
nitrogen and total phosphorus from soil leachate in subsurface wetland
microcosms. Bioresource Technology 94:185-192.
Freeland, R. O. 1944. Apparent photosynthesis in some conifers during the winter. Plant
Physiology 19: 179-185.
Gagnon V., F. Chazarenc, Y. Comeau, and J. Brisson 2006. Influence of macrophytes
species on microbial density and activity in constructed wetlands. Pages 10251033 In 10th International Conference on Wetland Systems for Water Pollution
Control. Pergamon Press. Oxford, UK
Gersberg, R. M., B. V. Elkins, S. R. Lyon, and C. R. Goldman. 1986. Role of aquatic
plants in waste-water treatment by artificial wetlands. Water Research 20:363368.
Gries, C., L. Kappen, and R. Losch. 1990. Mechanism of flood tolerance in reed,
Phragmites australis (Cav) Trin Ex Steudel. New Phytologist 114:589-593.
Griffin, P., P. Jennings, and E. Bowman. 1999. Advanced nitrogen removal by rotating
biological contactors, recycle and constructed wetlands. Water Science and
Technology 40:383-390.
Haberl R. and Perfler R. 1990. Seven years of research work and experience with
wastewater treatment by a reed bed system. Pages 41-51 In International
Conference on the Use of Constructed Wetlands in Water Pollution Control.
Pergamon Press. Oxford, UK
76
Hatano K., C.C. Trettin, C.H. House, and I. A. G. Wollum. 1993. Microbial populations
and decomposition activity in three subsurface flow constructed wetlands. Pages
541-547 in G. A. Moshiri. editor. Constructed wetland for water quality
improvement. Lewis Publishers, London.
Hook, P. B., O. R. Stein, W. C. Allen, J. A. Biederman, 2003. Plant species effects on
seasonal performance patterns in model subsurface wetlands. Pages 87-106 In
Mander, U. and P. D. Jenssen. editor. Constructed Wetlands for Wastewater
Treatment in Cold Climates. WIT Press, Southampton
Howes, B. L., and J. M. Teal. 1994. Oxygen loss from Spartina alterniflora and its
relationship to salt-marsh oxygen balance. Oecologia 97:431-438.
Hunter, R. G., D. L. Combs, and D. B. George. 2001. Nitrogen, phosphorous, and organic
carbon removal in simulated wetland treatment systems. Archives of
Environmental Contamination and Toxicology 41:274-281.
Iamchaturapatr, J., S. W. Yi, and J. S. Rhee. 2007. Nutrient removals by 21 aquatic plants
for vertical free surface-flow (VFS) constructed wetland. Ecological Engineering
29:287-293.
Jackson, M. B., and W. Armstrong. 1999. Formation of aerenchyma and the processes of
plant ventilation in relation to soil flooding and submergence. Plant Biology
1:274-287.
Jackson M.B. and M.C. Drew. 1984. Effects of flooding on growth and metabolism of
herbaceous plants. Pages 47-128 In T.T. Kozlowski, Flooding and Plant Growth.
Academic Press Inc., New York, NY
Justin, S. H. F. W., and W. Armstrong. 1987. The anatomical characteristics of roots and
plant-response to soil flooding. New Phytologist 106:465-495.
Kadlec R. H., and R. L. Knight. 1996. Treatment Wetlands. Lewis Publishers, Boca
Raton, FL.
Kadlec, R. H., and K. R. Reddy. 2001. Temperature effects in treatment wetlands. Water
Environment Research 73:543-557.
Kludze, H. K., and R. D. Delaune. 1994. Methane emissions and growth of Spartina
patens in response to soil redox intensity. Soil Science Society of America
Journal 58:1838-1845.
Kludze, H. K., R. D. Delaune, and W. H. Patrick. 1994. A colorimetric method for
assaying dissolved-oxygen loss from container-grown rice roots. Agronomy
Journal 86:483-487.
77
Koncalova, H. 1990. Anatomical adaptations to waterlogging in roots of wetland
graminoids - limitations and drawbacks. Aquatic Botany 38:127-134.
Kyambadde, J., F. Kansiime, L. Gumaelius, and G. Dalhammar. 2004. A comparative
study of Cyperus papyrus and Miscanthidium violaceum-based constructed
wetlands for wastewater treatment in a tropical climate. Water Research 38:475485.
Laan, P., M. J. Berrevoets, S. Lythe, W. Armstrong, and C. W. P. M. Blom. 1989. Root
morphology and aerenchyma formation as indicators of the flood-tolerance of
Rumex species. Journal of Ecology 77:693-703.
Laskov, C., O. Horn, and M. Hupfer. 2006. Environmental factors regulating the radial
oxygen loss from roots of Myriophyllum spicatum and Potamogeton crispus.
Aquatic Botany 84:333-340.
Lin, Y. F., S. R. Jing, T. W. Wang, and D. Y. Lee. 2002. Effects of macrophytes and
external carbon sources on nitrate removal from groundwater in constructed
wetlands. Environmental Pollution 119:413-420.
Luxmore, R. J., L. H. Stolzy, and J. Letey. 1970. Oxygen diffusion in the soil-plant
system. Agronomy Journal 62:317-332.
Lynch, J. M., and J. M. Whipps. 1990. Substrate flow in the rhizosphere. Plant and Soil
129:1-10.
Machemer, S. D., and T. R. Wildeman. 1992. Adsorption compared with sulfide
precipitation as metal removal processes from acid-mine drainage in a constructed
wetland. Journal of Contaminant Hydrology 9:115-131.
Mander, U. and P. D. Jenssen. 2003. Constructed Wetlands for Wastewater Treatment in
Cold Climates. WIT Press, Southampton, UK
Matsui, T., and T. Tsuchiya. 2006. A method to estimate practical radial oxygen loss of
wetland plant roots. Plant and Soil 279: 119-128.
McManmon, M., and R. M. Crawford. 1971. Metabolic theory of flooding tolerance significance of enzyme distribution and behaviour. New Phytologist 70:299-306.
Moog, P. R. 1998. Flooding tolerance of Carex species. I. Root structure. Planta 207:189198.
Moog, P. R., and W. Bruggemann. 1998. Flooding tolerance of Carex species. II. Root
gas-exchange capacity. Planta 207:199-206.
78
Munch, C., P. Kuschk, and I. Roske. 2005. Root stimulated nitrogen removal: only a
local effect or important for water treatment? Water Science and Technology
51:185-192.
Nelson, E. A., W. L. Specht, and A. S. Knox. 2006. Metal removal from water discharges
by a constructed treatment wetland. Engineering in Life Sciences 6:26-30.
Nivala, J., M. B. Hoos, C. Cross, S. Wallace, and G. Parkin. 2007. Treatment of landfill
leachate using an aerated, horizontal subsurface-flow constructed wetland.
Science of the Total Environment 380:19-27.
Ottova, V., J. Balcarova, and J. Vymazal. 1997. Microbial characteristics of constructed
wetlands. Water Science and Technology 35:117-123.
Parker, J. and E. Winkler. 1953. Photosynthesis of Picea excelsa in winter. Ecology 34:
605-609.
Picard, C. R., L. H. Fraser, and D. Steer. 2005. The interacting effects of temperature and
plant community type on nutrient removal in wetland microcosms. Bioresource
Technology 96:1039-1047.
Ponnamperuma, F.N. 1972. The chemistry of submerged soils. Advances in Agronomy.
24:29-96.
Reddy, K. R., E. M. D'Angelo, and T. A. DeBusk. 1989. Oxygen transport through
aquatic macrophytes: the role in wastewater treatment. Journal of Environmental
Quality 19:261-267.
Reddy, K. R., W. H. Patrick, and C. W. Lindau. 1989. Nitrification-denitrification at the
plant root-sediment interface in wetlands. Limnology and Oceanography 34:10041013.
Reed, S. C., and D. S. Brown. 1992. Constructed wetland design - the First Generation.
Water Environment Research 64:776-781.
Schluter, U. and R. M. M. Crawford. 2001. Long-term anoxia tolerance in leaves of
Acorus calamus and Iris pseudacorus. Experimental Botany 52:2213-2225.
Sorrell, B. K. 1999. Effect of external oxygen demand on radial oxygen loss by Juncus
roots in titanium citrate solutions. Plant Cell and Environment 22:1587-1593.
Sorrell, B. K., and W. Armstrong. 1994. On the difficulties of measuring oxygen release
by root systems of wetland plants. Journal of Ecology 82:177-183.
79
Sorrell, B. K., I. A. Mendelssohn, K. L. Mckee, and R. A. Woods. 2000. Ecophysiology
of wetland plant roots: A modeling comparison of aeration in relation to species
distribution. Annals of Botany 86:675-685.
Stein, O. R., and P. B. Hook. 2005. Temperature, plants, and oxygen: How does season
affect constructed wetland performance? Journal of Environmental Science and
Health 40:1331-1342.
Stein, O. R., D.J. Borden-Stewart, P.B. Hook and W. L. Jones. 2007. Seasonal influences
on sulfate reduction and metal sequestration in sub-surface wetlands. Water
Research 41: 3440-3448.
Tanaka, N., K. Yutani, T. Aye, and K. B. S. N. Jinadasa. 2007. Effect of broken dead
culms of Phragmites australis on radial oxygen loss in relation to radiation and
temperature. Hydrobiologia 583:165-172.
Tanner, C. C. 2001. Plants as ecosystem engineers in subsurface-flow treatment wetlands.
Water Science and Technology 44:9-17.
Tolley, M. D., R. D. Delaune, and W. H. Patrick. 1986. The effect of sediment redox
potential and soil acidity on nitrogen uptake, anaerobic root respiration, and
growth of rice (Oryza sativa). Plant and Soil 93:323-331.
USEPA. 1993. Subsurface Flow Constructed Wetlands for Wastewater Treatment: A
Technology Assessment. EPA/832/R/93/001. USEPA, Office of Water,
Washington, DC.
USEPA. 2000. Manual: Constructed Wetlands Treatment of Municipal Wastewaters.
EPA/625/R-99//010. USEPA, National Risk Management Research Laboratory,
Office and Research Development, Cincinnati, OH.
U.S. Fish and Wildlife Service. 1988. National list of vascular plant species that occur in
wetlands. U.S. Fish and Wildlife Service Biological Report 88. Department of the
Interior, Washington, D.C.
Vacca, G., H. Wand, M. Nikolausz, P. Kuschk, and M. Kastner. 2005. Effect of plants
and filter materials on bacteria removal in pilot-scale constructed wetlands. Water
Research 39:1361-1373.
Veneman, P. L. M., and E. W. Pickering. 1983. Salt bridge for field redox potential
measurements. Communications in Soil Science and Plant Analysis 14:669-677.
Vepraskas, M. J., J. L. Cox, and students of the Soil Science Dept. 2002. NC State
University, Raleigh, NC
80
Visser, E. J. W., C. W. P. M. Blom, and L. A. C. J. Voesenek. 1996. Flooding-induced
adventitious rooting in Rumex: Morphology and development in an ecological
perspective. Acta Botanica Nederlandica 45:17-28.
Visser, E. J. W., and G. M. Bogemann. 2003. Measurement of porosity in very small
samples of plant tissue. Plant and Soil 253:81-90.
Visser, E. J. W., G. M. Bogemann, H. M. Van de Steeg, R. Pierik, and C. W. P. M. Blom.
2000. Flooding tolerance of Carex species in relation to field distribution and
aerenchyma formation. New Phytologist 148:93-103.
Visser, E. J. W., J. D. Cohen, G. W. M. Barendse, C. W. P. M. Blom, and L. A. C. J.
Voesenek. 1996. An ethylene-mediated increase in sensitivity to auxin induces
adventitious root formation in flooded Rumex palustris Sm. Plant Physiology
112:1687-1692.
Visser, E. J. W., T. D. Colmer, C. W. P. M. Blom, and L. A. C. J. Voesenek. 2000.
Changes in growth, porosity, and radial oxygen loss from adventitious roots of
selected mono- and dicotyledonous wetland species with contrasting types of
aerenchyma. Plant Cell and Environment 23:1237-1245.
Weiss, J. V., D. Emerson, S. M. Backer, and J. P. Megonigal. 2003. Enumeration of
Fe(II)-oxidizing and Fe(III)-reducing bacteria in the root zone of wetland plants:
Implications for a rhizosphere iron cycle. Biogeochemistry 64:77-96.
Yang, Q., Z. H. Chen, J. G. Zhao, and B. H. Gu. 2007. Contaminant removal of domestic
wastewater by constructed wetlands: Effects of plant species. Journal of
Integrative Plant Biology 49:437-446.
Zehnder, A. J. B., and K. Wuhrmann. 1976. Titanium(III) citrate as a nontoxic oxidationreduction buffering system for culture of obligate anaerobes. Science 194:11651166.